Download to read offline
![v
Preface
This book discusses the multiple facets of “Bioinformatics for Omics Data,” an area of
research that intersects with and integrates diverse disciplines, including molecular biol-
ogy, applied informatics, and statistics, among others. Bioinformatics has become a default
technology for data-driven research in the Omics realm and a necessary skill set for the
Omics practitioner. Progress in miniaturization, coupled with advancements in readout
technologies, has enabled a multitude of cellular components and states to be assessed
simultaneously, providing an unparalleled ability to characterize a given biological pheno-
type. However, without appropriate processing and analysis, Omics data add nothing to
our understanding of the phenotype under study. Even managing the enormous amounts
of raw data that these methods generate has become something of an art.
Viewed from one perspective, bioinformatics might be perceived as a purely technical
discipline. However, as a research discipline, bioinformatics might more accurately be viewed
as “[molecular] biology involving computation.” Omics has triggered a paradigm shift in
experimental study design, expanding beyond hypothesis-driven approaches to research that
is basically explorative. At present, Omics is in the process of consolidating various interme-
diate forms between these two extremes. In this context, bioinformatics for Omics data
serves both hypothesis generation and validation and is thus much more than mere data
management and processing. Bioinformatics workflows with data interpretation strategies
that reflect the complexity of biological organization have been designed. These approaches
interrogate abundance profiles with regulatory elements, all expressed as interaction net-
works, thus allowing a one-step (descriptive) embodiment of wide-ranging cellular pro-
cesses. Here, the seamless transition to computational Systems Biology becomes apparent,
the ultimate goal of which is representing the dynamics of a phenotype in quantitative mod-
els capable of predicting the emergence of higher order molecular procedures and functions
that arise from the interplay of basic molecular entities that constitute a living cell.
Bioinformatics for Omics data is certainly embedded in a highly complex technologi-
cal and scientific environment, but it is also a component and driver of one of the most
exciting developments in modern molecular biology. Thus, while this book seeks to pro-
vide practical guidelines, it hopefully also conveys a sense of fascination associated with
this research field.
This volume is structured in three parts. Part I provides central analysis strategies,
standardization, and data management guidelines, as well as fundamental statistics for
analyzing Omics profiles. Part II addresses bioinformatics approaches for specific Omics
tracks, spanning genome, transcriptome, proteome, and metabolome levels. For each
track, the conceptual and experimental background is provided, together with specific
guidelines for handling raw data, including preprocessing and analysis. Part III presents
examples of integrated Omics bioinformatics applications, complemented by case studies
on biomarker and target identification in the context of human disease.
I wish to express my gratitude to all authors for their dedication in providing excellent
chapters, and to John Walker, who initiated this project. As for any omissions or errors,
the responsibility is mine. In any case, enjoy reading.
Vienna, Austria Bernd Mayer](https://image.slidesharecdn.com/13698-250224053558-3b158fc4/85/Bioinformatics-for-Omics-Data-Methods-and-Protocols-1st-Edition-Maria-V-Schneider-11-320.jpg)
Bioinformatics for Omics Data Methods and Protocols 1st Edition Maria V. Schneider
- 1. Visit https://ebookultra.com todownload the full version and explore more ebooks Bioinformatics for Omics Data Methods and Protocols 1st Edition Maria V. Schneider _____ Click the link below to download _____ https://ebookultra.com/download/bioinformatics-for- omics-data-methods-and-protocols-1st-edition-maria-v- schneider/ Explore and download more ebooks at ebookultra.com
- 2. Here are somesuggested products you might be interested in. Click the link to download Plant Bioinformatics Methods and Protocols 3rd Edition David Edwards https://ebookultra.com/download/plant-bioinformatics-methods-and- protocols-3rd-edition-david-edwards/ Cell Penetrating Peptides Methods and Protocols 1st Edition Maria Lindgren https://ebookultra.com/download/cell-penetrating-peptides-methods-and- protocols-1st-edition-maria-lindgren/ Statistical Methods for Microarray Data Analysis Methods and Protocols 1st Edition Andrei Y. Yakovlev https://ebookultra.com/download/statistical-methods-for-microarray- data-analysis-methods-and-protocols-1st-edition-andrei-y-yakovlev/ Integrating Omics Data 1st Edition George C. Tseng https://ebookultra.com/download/integrating-omics-data-1st-edition- george-c-tseng/
- 3. Data Mining forBioinformatics Applications 1st Edition He Zengyou https://ebookultra.com/download/data-mining-for-bioinformatics- applications-1st-edition-he-zengyou/ Java Methods Object Oriented Programming and Data Structures Third AP 3rd AP Edition Maria Litvin https://ebookultra.com/download/java-methods-object-oriented- programming-and-data-structures-third-ap-3rd-ap-edition-maria-litvin/ Ubiquitin Proteasome Protocols 1st Edition Maria Gaczynska https://ebookultra.com/download/ubiquitin-proteasome-protocols-1st- edition-maria-gaczynska/ Feature Selection and Ensemble Methods for Bioinformatics Algorithmic Classification and Implementations 1st Edition Oleg Okun https://ebookultra.com/download/feature-selection-and-ensemble- methods-for-bioinformatics-algorithmic-classification-and- implementations-1st-edition-oleg-okun/ Data Mining Multimedia Soft Computing and Bioinformatics 1st Edition Sushmita Mitra https://ebookultra.com/download/data-mining-multimedia-soft-computing- and-bioinformatics-1st-edition-sushmita-mitra/
- 5. Bioinformatics for OmicsData Methods and Protocols 1st Edition Maria V. Schneider Digital Instant Download Author(s): Maria V. Schneider, Sandra Orchard (auth.), Bernd Mayer (eds.) ISBN(s): 9781617790263, 1617790265 Edition: 1 File Details: PDF, 39.09 MB Year: 2011 Language: english
- 7. Me t ho d s i n Mo l e c u l a r Bi o l o g y ™ Series Editor John M. Walker School of Life Sciences University of Hertfordshire Hatfield, Hertfordshire, AL10 9AB, UK For other titles published in this series, go to www.springer.com/series/7651
- 9. Bioinformatics for OmicsData Methods and Protocols Edited by Bernd Mayer emergentecbiodevelopmentGmbH,Vienna,Austria
- 10. Editor Bernd Mayer, Ph.D. emergentecbiodevelopment GmbH Gersthofer Strasse 29-31 1180 Vienna, Austria bernd.mayer@emergentec.com and Institute for Theoretical Chemistry University of Vienna Währinger Strasse 17 1090 Vienna, Austria bernd.mayer@univie.ac.at ISSN 1064-3745 e-ISSN 1940-6029 ISBN 978-1-61779-026-3 e-ISBN 978-1-61779-027-0 DOI 10.1007/978-1-61779-027-0 Springer New York Dordrecht Heidelberg London Library of Congress Control Number: 2011922257 © Springer Science+Business Media, LLC 2011 All rights reserved. This work may not be translated or copied in whole or in part without the written permission of the publisher (Humana Press, c/o Springer Science+Business Media, LLC, 233 Spring Street, New York, NY 10013, USA), except for brief excerpts in connection with reviews or scholarly analysis. Use in connection with any form of information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed is forbidden. The use in this publication of trade names, trademarks, service marks, and similar terms, even if they are not identified as such, is not to be taken as an expression of opinion as to whether or not they are subject to proprietary rights. While the advice and information in this book are believed to be true and accurate at the date of going to press, neither the authors nor the editors nor the publisher can accept any legal responsibility for any errors or omissions that may be made. The publisher makes no warranty, express or implied, with respect to the material contained herein. Printed on acid-free paper Humana Press is part of Springer Science+Business Media (www.springer.com)
- 11. v Preface This book discussesthe multiple facets of “Bioinformatics for Omics Data,” an area of research that intersects with and integrates diverse disciplines, including molecular biol- ogy, applied informatics, and statistics, among others. Bioinformatics has become a default technology for data-driven research in the Omics realm and a necessary skill set for the Omics practitioner. Progress in miniaturization, coupled with advancements in readout technologies, has enabled a multitude of cellular components and states to be assessed simultaneously, providing an unparalleled ability to characterize a given biological pheno- type. However, without appropriate processing and analysis, Omics data add nothing to our understanding of the phenotype under study. Even managing the enormous amounts of raw data that these methods generate has become something of an art. Viewed from one perspective, bioinformatics might be perceived as a purely technical discipline. However, as a research discipline, bioinformatics might more accurately be viewed as “[molecular] biology involving computation.” Omics has triggered a paradigm shift in experimental study design, expanding beyond hypothesis-driven approaches to research that is basically explorative. At present, Omics is in the process of consolidating various interme- diate forms between these two extremes. In this context, bioinformatics for Omics data serves both hypothesis generation and validation and is thus much more than mere data management and processing. Bioinformatics workflows with data interpretation strategies that reflect the complexity of biological organization have been designed. These approaches interrogate abundance profiles with regulatory elements, all expressed as interaction net- works, thus allowing a one-step (descriptive) embodiment of wide-ranging cellular pro- cesses. Here, the seamless transition to computational Systems Biology becomes apparent, the ultimate goal of which is representing the dynamics of a phenotype in quantitative mod- els capable of predicting the emergence of higher order molecular procedures and functions that arise from the interplay of basic molecular entities that constitute a living cell. Bioinformatics for Omics data is certainly embedded in a highly complex technologi- cal and scientific environment, but it is also a component and driver of one of the most exciting developments in modern molecular biology. Thus, while this book seeks to pro- vide practical guidelines, it hopefully also conveys a sense of fascination associated with this research field. This volume is structured in three parts. Part I provides central analysis strategies, standardization, and data management guidelines, as well as fundamental statistics for analyzing Omics profiles. Part II addresses bioinformatics approaches for specific Omics tracks, spanning genome, transcriptome, proteome, and metabolome levels. For each track, the conceptual and experimental background is provided, together with specific guidelines for handling raw data, including preprocessing and analysis. Part III presents examples of integrated Omics bioinformatics applications, complemented by case studies on biomarker and target identification in the context of human disease. I wish to express my gratitude to all authors for their dedication in providing excellent chapters, and to John Walker, who initiated this project. As for any omissions or errors, the responsibility is mine. In any case, enjoy reading. Vienna, Austria Bernd Mayer
- 13. vii Contents Preface . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v Contributors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix Part I Omics Bioinformatics Fundamentals 1 Omics Technologies, Data and Bioinformatics Principles . . . . . . . . . . . . . . . . . . . 3 Maria V. Schneider and Sandra Orchard 2 Data Standards for Omics Data: The Basis of Data Sharing and Reuse . . . . . . . . . 31 Stephen A. Chervitz, Eric W. Deutsch, Dawn Field, Helen Parkinson, John Quackenbush, Phillipe Rocca-Serra, Susanna-Assunta Sansone, Christian J. Stoeckert, Jr., Chris F. Taylor, Ronald Taylor, and Catherine A. Ball 3 Omics Data Management and Annotation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71 Arye Harel, Irina Dalah, Shmuel Pietrokovski, Marilyn Safran, and Doron Lancet 4 Data and Knowledge Management in Cross-Omics Research Projects . . . . . . . . . 97 Martin Wiesinger, Martin Haiduk, Marco Behr, Henrique Lopes de Abreu Madeira, Gernot Glöckler, Paul Perco, and Arno Lukas 5 Statistical Analysis Principles for Omics Data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Daniela Dunkler, Fátima Sánchez-Cabo, and Georg Heinze 6 Statistical Methods and Models for Bridging Omics Data Levels . . . . . . . . . . . . . 133 Simon Rogers 7 Analysis of Time Course Omics Datasets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153 Martin G. Grigorov 8 The Use and Abuse of -Omes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173 Sonja J. Prohaska and Peter F. Stadler Part II Omics Data and Analysis Tracks 9 Computational Analysis of High Throughput Sequencing Data . . . . . . . . . . . . . . 199 Steve Hoffmann 10 Analysis of Single Nucleotide Polymorphisms in Case–Control Studies . . . . . . . . 219 Yonghong Li, Dov Shiffman, and Rainer Oberbauer 11 Bioinformatics for Copy Number Variation Data . . . . . . . . . . . . . . . . . . . . . . . . . 235 Melissa Warden, Roger Pique-Regi, Antonio Ortega, and Shahab Asgharzadeh 12 Processing ChIP-Chip Data: From the Scanner to the Browser . . . . . . . . . . . . . . 251 Pierre Cauchy, Touati Benoukraf, and Pierre Ferrier 13 Insights Into Global Mechanisms and Disease by Gene Expression Profiling . . . . 269 Fátima Sánchez-Cabo, Johannes Rainer, Ana Dopazo, Zlatko Trajanoski, and Hubert Hackl
- 14. viii Contents 14 Bioinformaticsfor RNomics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 299 Kristin Reiche, Katharina Schutt, Kerstin Boll, Friedemann Horn, and Jörg Hackermüller 15 Bioinformatics for Qualitative and Quantitative Proteomics . . . . . . . . . . . . . . . . . 331 Chris Bielow, Clemens Gröpl, Oliver Kohlbacher, and Knut Reinert 16 Bioinformatics for Mass Spectrometry-Based Metabolomics . . . . . . . . . . . . . . . . . 351 David P. Enot, Bernd Haas, and Klaus M. Weinberger Part III Applied Omics Bioinformatics 17 Computational Analysis Workflows for Omics Data Interpretation . . . . . . . . . . . . 379 Irmgard Mühlberger, Julia Wilflingseder, Andreas Bernthaler, Raul Fechete, Arno Lukas, and Paul Perco 18 Integration, Warehousing, and Analysis Strategies of Omics Data . . . . . . . . . . . . . 399 Srinubabu Gedela 19 Integrating Omics Data for Signaling Pathways, Interactome Reconstruction, and Functional Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 415 Paolo Tieri, Alberto de la Fuente, Alberto Termanini, and Claudio Franceschi 20 Network Inference from Time-Dependent Omics Data . . . . . . . . . . . . . . . . . . . . 435 Paola Lecca, Thanh-Phuong Nguyen, Corrado Priami, and Paola Quaglia 21 Omics and Literature Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457 Vinod Kumar 22 Omics–Bioinformatics in the Context of Clinical Data . . . . . . . . . . . . . . . . . . . . . 479 Gert Mayer, Georg Heinze, Harald Mischak, Merel E. Hellemons, Hiddo J. Lambers Heerspink, Stephan J.L. Bakker, Dick de Zeeuw, Martin Haiduk, Peter Rossing, and Rainer Oberbauer 23 Omics-Based Identification of Pathophysiological Processes . . . . . . . . . . . . . . . . . 499 Hiroshi Tanaka and Soichi Ogishima 24 Data Mining Methods in Omics-Based Biomarker Discovery . . . . . . . . . . . . . . . . 511 Fan Zhang and Jake Y. Chen 25 Integrated Bioinformatics Analysis for Cancer Target Identification . . . . . . . . . . . 527 Yongliang Yang, S. James Adelstein, and Amin I. Kassis 26 Omics-Based Molecular Target and Biomarker Identification . . . . . . . . . . . . . . . . 547 Zgang–Zhi Hu, Hongzhan Huang, Cathy H. Wu, Mira Jung, Anatoly Dritschilo, Anna T. Riegel, and Anton Wellstein Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
- 15. ix Contributors S. James Adelstein• Harvard Medical School, Harvard University, Boston, MA, USA Shahab Asgharzadeh • Department of Pediatrics and Pathology, Keck School of Medicine, Childrens Hospital Los Angeles, University of Southern California, Los Angeles, CA, USA Stephan J.L. Bakker • Department of Nephrology, University Medical Center Groningen, Groningen, The Netherlands Catherine A. Ball • Department of Genetics, Stanford University School of Medicine, Stanford, CA, USA Marco Behr • emergentec biodevelopment GmbH, Vienna, Austria Touati Benoukraf • Université de la Méditerranée, Marseille, France; Centre d’Immunologie de Marseille-Luminy, Marseille, France; CNRS, UMR6102, Marseille, France; Inserm, U631, Marseille, France Andreas Bernthaler • emergentec biodevelopment GmbH, Vienna, Austria Chris Bielow • AG Algorithmische Bioinformatik, Institut für Informatik, Freie Universität Berlin, Berlin, Germany Pierre Cauchy • Inserm, U928, TAGC, Marseille, France; Université de la Méditerranée, Marseille, France Jake Y. Chen • Indiana University School of Informatics, Indianapolis, IN, USA Stephen A. Chervitz • Affymetrix, Inc., Santa Clara, CA, USA Irina Dalah • Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel Eric W. Deutsch • Institute for Systems Biology, Seattle,WA, USA Ana Dopazo • Genomics Unit, Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain Anatoly Dritschilo • Lombardi Cancer Center, Georgetown University, Washington, DC, USA Daniela Dunkler • Section of Clinical Biometrics, Center for Medical Statistics, Informatics and Intelligent Systems, Medical University of Vienna, Vienna, Austria David P. Enot • BIOCRATES life sciences AG, Innsbruck, Austria Raul Fechete • emergentec biodevelopment GmbH, Vienna, Austria Pierre Ferrier • Centre d’Immunologie de Marseille-Luminy (CIML), Marseille, France Dawn Field • NERC Centre for Ecology and Hydrology, Oxford, UK Claudio Franceschi • ‘L Galvani’ Interdept Center, University of Bologna, Bologna, Italy Alberto de la Fuente • CRS4 Bioinformatica, Parco Tecnologico SOLARIS, Pula, Italy Srinubabu Gedela • Stanford University School of Medicine, Stanford, CA, USA Gernot Glöckler • emergentec biodevelopment GmbH, Vienna, Austria Martin G. Grigorov • Nestlé Research Center, Lausanne, Switzerland Clemens Gröpl • Ernst-Moritz-Arndt-Universität Greifswald, Greifswald, Germany Bernd Haas • BIOCRATES life sciences AG, Innsbruck, Austria
- 16. x Contributors Jörg Hackermüller• Bioinformatics Group, Department of Computer Science, University of Leipzig, Leipzig, Germany; Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany Hubert Hackl • Division for Bioinformatics, Innsbruck Medical University, Innsbruck, Austria Martin Haiduk • emergentec biodevelopment GmbH, Vienna, Austria Arye Harel • Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel Georg Heinze • Section of Clinical Biometrics, Center for Medical Statistics, Informatics and Intelligent Systems, Medical University of Vienna, Vienna, Austria Merel E. Hellemons • Department of Nephrology, University Medical Center Groningen, Groningen, The Netherlands Steve Hoffmann • Interdisciplinary Center for Bioinformatics and The Junior Research Group for Transcriptome Bioinformatics in the LIFE Research Cluster, University Leipzig, Leipzig, Germany Friedemann Horn • Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany; Institute of Clinical Immunology, University of Leipzig, Leipzig, Germany Zhang-Zhi Hu • Lombardi Cancer Center, Georgetown University, Washington DC, USA Hongzhan Huang • Center for Bioinformatics Computational Biology, University of Delaware, Newark, DE, USA Mira Jung • Lombardi Cancer Center, Georgetown University, Washington, DC, USA Amin I. Kassis • Harvard Medical School, Harvard University, Boston, MA, USA Oliver Kohlbacher • Eberhard-Karls-Universität Tübingen, Tübingen, Germany Vinod Kumar • Computational Biology, Quantitative Sciences, GlaxoSmithKline, King of Prussia, PA, USA Hiddo J. Lambers Heerspink • Department of Nephrology, University Medical Center Groningen, Groningen, The Netherlands Doron Lancet • Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel Paola Lecca • The Microsoft Research – University of Trento Centre for Computational and Systems Biology, Povo, Trento, Italy Yonghong Li • Celera Corporation, Alameda, CA, USA Henrique Lopes de Abreu Madeira • emergentec biodevelopment GmbH, Vienna, Austria Arno Lukas • emergentec biodevelopment GmbH, Vienna, Austria Gert Mayer • Department of Internal Medicine IV (Nephrology and Hypertension), Medical University of Innsbruck, Innsbruck, Austria Harald Mischak • mosaiques diagnostics GmbH, Hannover, Germany Irmgard Mühlberger • emergentec biodevelopment GmbH, Vienna, Austria Thanh-Phuong Nguyen • The Microsoft Research – University of Trento Centre for Computational and Systems Biology, Povo, Trento, Italy Rainer Oberbauer • Medical University of Vienna and KH Elisabethinen Linz, Vienna, Austria Soichi Ogishima • Department of Bioinformatics, Medical Research Institute, Tokyo Medical and Dental University, Tokyo, Japan
- 17. xi Contributors Sandra Orchard •EMBL-European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton,Cambridge, UK Antonio Ortega • Department of Electrical Engineering, Viterbi School of Engineering, University of Southern California, Los Angeles, CA, USA Helen Parkinson • EMBL-EBI, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK Paul Perco • emergentec biodevelopment GmbH, Vienna, Austria Shmuel Pietrokovski • Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel Roger Pique-Regi • Department of Human Genetics, University of Chicago, Chicago, IL, USA Corrado Priami • The Microsoft Research – University of Trento Centre for Computational and Systems Biology, Povo, Trento, Italy Sonja J. Prohaska • Department of Computer Science and Interdisciplinary Center for Bioinformatics, University of Leipzig, Leipzig, Germany John Quackenbush • Department of Biostatistics, Dana-Farber Cancer Institute, Boston, MA, USA Paola Quaglia • The Microsoft Research – University of Trento Centre for Computational and Systems Biology, Povo, Trento, Italy Johannes Rainer • Bioinformatics Group, Division Molecular Pathophysiology, Medical University Innsbruck, Innsbruck, Austria Kristin Reiche • Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany Knut Reinert • AG Algorithmische Bioinformatik, Institut für Informatik, Freie Universität Berlin, Berlin, Germany Anna T. Riegel • Lombardi Cancer Center, Georgetown University, Washington, DC, USA Phillipe Rocca-Serra • EMBL-EBI, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK Simon Rogers • Inference Research Group, Department of Computing Science, University of Glasgow, Glasgow, UK Peter Rossing • Steno Diabetes Center Denmark, Gentofte, Denmark Marilyn Safran • Department of Molecular Genetics, Weizmann Institute of Science, Rehovot, Israel Fátima Sánchez-Cabo • Genomics Unit, Centro Nacional de Investigaciones Cardiovasculares, Madrid, Spain Susanna-Assunta Sansone • EMBL-EBI, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK Maria V. Schneider • EMBL-European Bioinformatics Institute, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK Katharina Schutt • Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany; Institute of Clinical Immunology, University of Leipzig, Leipzig, Germany Dov Shiffman • Celera Corporation, Alameda, CA, USA Peter F. Stadler • Department of Computer Science and Interdisciplinary Center for Bioinformatics, University of Leipzig, Leipzig, Germany
- 18. xii Contributors Christian J.Stoeckert Jr • Department of Genetics and Center for Bioinformatics, University of Pennsylvania School of Medicine, Philadelphia, PA, USA Hiroshi Tanaka • Department of Computational Biology, Graduate School of Biomedical Science, Tokyo Medical and Dental University, Tokyo, Japan Chris F. Taylor • EMBL-EBI, Wellcome Trust Genome Campus, Hinxton, Cambridge, UK Ronald Taylor • Computational Biology Bioinformatics Group, Pacific Northwest National Laboratory, Richland, WA, USA Alberto Termanini • ‘L Galvani’ Interdept Center, University of Bologna, Bologna, Italy Paolo Tieri • ‘L Galvani’ Interdept Center, University of Bologna, Bologna, Italy Zlatko Trajanoski • Division for Bioinformatics, Innsbruck Medical University, Innsbruck, Austria Kerstin Boll • Fraunhofer Institute for Cell Therapy and Immunology, Leipzig, Germany; Institute of Clinical Immunology, University of Leipzig, Leipzig, Germany Melissa Warden • Department of Pediatrics and Pathology, Keck School of Medicine, Childrens Hospital Los Angeles, University of Southern California, Los Angeles, CA, USA Klaus M. Weinberger • BIOCRATES life sciences AG, Innsbruck, Austria Anton Wellstein • Lombardi Cancer Center, Georgetown University, Washington, DC, USA Martin Wiesinger • emergentec biodevelopment GmbH, Vienna, Austria Julia Wilflingseder • Medical University of Vienna and KH Elisabethinen Linz, Vienna, Austria Cathy H. Wu • Center for Bioinformatics Computational Biology, University of Delaware, Newark, DE, USA Yongliang Yang • Department of Radiology, Harvard Medical School, Harvard University, Boston, MA, USA; Center of Molecular Medicine, Department of Biological Engineering, Dalian University of Technology, Dalian, China Dick de Zeeuw • Department of Nephrology, University Medical Center Groningen, Groningen, The Netherlands Fan Zhang • Indiana University School of Informatics, Indianapolis, IN, USA
- 20. Part I Omics BioinformaticsFundamentals
- 22. 3 Bernd Mayer (ed.),Bioinformatics for Omics Data: Methods and Protocols, Methods in Molecular Biology, vol. 719, DOI 10.1007/978-1-61779-027-0_1, © Springer Science+Business Media, LLC 2011 Chapter 1 Omics Technologies, Data and Bioinformatics Principles Maria V. Schneider and Sandra Orchard Abstract We provide an overview on the state of the art for the Omics technologies, the types of omics data and the bioinformatics resources relevant and related to Omics. We also illustrate the bioinformatics chal- lenges of dealing with high-throughput data. This overview touches several fundamental aspects of Omics and bioinformatics: data standardisation, data sharing, storing Omics data appropriately and exploring Omics data in bioinformatics. Though the principles and concepts presented are true for the various dif- ferent technological fields, we concentrate in three main Omics fields namely: genomics, transcriptomics and proteomics. Finally we address the integration of Omics data, and provide several useful links for bioinformatics and Omics. Key words: Omics, Bioinformatics, High-throughput, Genomics, Transcriptomics, Proteomics, Interactomics, Data integration, Omics databases, Omics tools The last decade has seen an explosion in the amount of biological data generated by an ever-increasing number of techniques enabling the simultaneous detection of a large number of altera- tions in molecular components (1). The Omics technologies uti- lise these high-throughput (HT) screening techniques to generate the large amounts of data required to enable a system level under- standing of correlations and dependencies between molecular components. Omics techniques are required to be high throughput because they need to analyse very large numbers of genes, gene expression, or proteins either in a single procedure or a combina- tion of procedures. Computational analysis, i.e., the discipline now known as bioinformatics, is a key requirement for the study of the vast amounts of data generated. Omics requires the use of 1. Introduction
- 23. 4 Schneider andOrchard techniques that can handle extremely complex biological samples in large quantities (e.g. high throughput) with high sensitivity and specificity. Next generation analytical tools require improved robustness, flexibility and cost efficiency. All of these aspects are being continuously improved, potentially enabling institutes such as the Wellcome Trust Sanger Sequencing Centre (see Note 1) to generate thousands of millions of base pairs per day, rather than the current output of 100 million per day (http:/ /www. yourgenome.org/sc/nt). However, all this data production makes sense only if one is equipped with the necessary analytical resources and tools to understand it. The evolution of the laboratory techniques has therefore to occur in parallel with a corresponding improvement in analytical methodology and tools to handle the data. The phrase Omics – a suffix signifying the measurement of the entire comple- ment of a given level of biological molecules and information – encompasses a variety of new technologies that can help explain both normal and abnormal cell pathways, networks, and processes via the simultaneous monitoring of thousands of molecular com- ponents. Bioinformaticians use computers and statistics to perform extensive Omics-related research by searching biological databases and comparing gene sequences and proteins on a vast scale to identify sequences or proteins that differ between diseased and healthy tissues, or more general between different phenotypes. “Omics” spans an increasingly wide range of fields, which now range from genomics (the quantitative study of protein coding genes, regulatory elements and noncoding sequences), transcrip- tomics (RNA and gene expression), proteomics (e.g. focusing on protein abundance), and metabolomics (metabolites and meta- bolic networks) to advances in the era of post-genomic biology and medicine: pharmacogenomics (the quantitative study of how genetics affects a host response to drugs), physiomics (physiologi- cal dynamics and functions of whole organisms) and in other fields: nutrigenomics (a rapidly growing discipline that focuses on iden- tifying the genetic factors that influence the body’s response to diet and studies how the bioactive constituents of food affect gene expression), phylogenomics (analysis involving genome data and evolutionary reconstructions, especially phylogenetics) and inter- actomics (molecular interaction networks). Though in the remain- der of this chapter we concentrate on an isolated few examples of Omics technologies, much of what is said, for example about data standardisation, data sharing, storage and analysis requirements are true for all of these different technological fields. There are already large amounts of data generated by these technologies and this trend is increasing, for example second and third generation sequencing technologies are leading to an exponential increase in the amount of sequencing data available. From a computational point of view, in order to address the
- 24. 5 Omics Technologies, Dataand Bioinformatics Principles complexity of these data, understand molecular regulation and gain the most from such comprehensive set of information, knowledge discovery – the process of automatically searching large volumes of data for patterns – is a crucial step. This process of bioinformatics analysis includes: (1) data processing and molecule (e.g. protein) identification, (2) statistical data analysis, (3) pathway analysis, and (4) data modelling in a system wide context. In this chapter we will present some of these analytical methods and discuss ways in which data can be made accessible to both the specialised bioinformatician, but in particular to the research scientist. There are a variety of definitions of the term HT; however we can loosely apply this term to cases where automation is used to increase the throughput of an experimental procedure. HT tech- nologies exploit robotics, optics, chemistry, biology and image analysis research. The explosion in data production in the public domain is a consequence of falling equipment prices, the opening of major national screening centres and new HT core facilities at universities and other academic institutes. The role of bioinfor- matics in HT technologies is of essential importance. High-Throughput Sequencing (HTS) technologies are used not only for traditional applications in genomics and metagenomics (see Note 2), but also for novel applications in the fields of tran- scriptomics, metatranscriptomics (see Note 3), epigenomics (see Note 4), and studies of genome variation (see Note 5). Next gen- eration sequencing platforms allow the determination of the sequence data from amplified single DNA fragments and have been developed specifically to lend themselves to robotics and par- allelisation. Current methods can directly sequence only relatively short (300–1,000 nucleotides long) DNA fragments in a single reaction. Short-read sequencing technologies dramatically reduce the sequencing cost. There were initial fears that the increase in quantity might result in a decrease in quality, and improvements in accuracy and read length are being looked for. However, despite this, these advances have significantly reduced the cost of several sequencing applications, such as resequencing individual genomes (2) readout assays (e.g. ChIP-seq (3) and RNAseq (4)). The transcriptome is the set of all messenger RNA (mRNA) mol- ecules, or “transcripts”, produced in one or a population of cells. Several methods have been developed in order to gain expression information at high throughput level. 2. Materials 2.1. Genomics High-Throughput Technologies 2.2. Transcriptomics High-Throughput Technologies
- 25. 6 Schneider andOrchard Global gene expression analysis has been conducted either by hybridization with oligonucleotide microarrays, or by counting of sequence tags. Digital transcriptomics with pyrophosphatase based ultra-high throughput DNA sequencing of ditags repre- sents a revolutionary approach to expression analysis, which gen- erates genome-wide expression profiles. ChIP-Seq is a technique that combines chromatin immunoprecipitation with sequencing technology to identify and quantify in vivo protein–DNA interac- tions on a genome-wide scale. Many of these applications are directly comparable to microarray experiments, for example ChIP-chip and ChIP-Seq are for all intents and purposes the same (5). The most recent increase in data generation in this evolving field is due to novel cycle-array sequencing methods (see Note 6), also known as next-generation sequencing (NGS), more com- monly described as second-generation sequencing which are already being used by technologies such as next-generation expressed-sequence-tag sequencing (see Note 7). Proteomics is the large-scale study of proteins, particularly their expression patterns, structures and functions, and there are vari- ous HT techniques applied to this area. Here we explore two main proteomics fields: Mass Spectrometry HT and Protein– Protein Interactions (PPIs). Mass spectrometry is an important emerging method for the characterization of proteins. It is also a rapidly developing field which is currently moving towards large-scale quantification of specific proteins in particular cell types under defined conditions. The rise of gel-free protein separation techniques, coupled with advances in MS instrumentation sensitivity and automation, has provided a foundation for high throughput approaches to the study of proteins. The identification of parent proteins from derived peptides now relies almost entirely on the software of search engines, which can perform in silico digests of protein sequence to generate peptides. Their molecular mass is then matched to the mass of the experimentally derived protein fragments. Studying protein–protein interactions provides valuable insights into many fields by helping precisely understand a protein’s role inside a specific cell type, with many of the techniques commonly used to experimentally determine protein interactions lending themselves to high throughput methodologies. Complementation assays (e.g. 2-hybrid) measure the oligomerisation-assisted com- plementation of two fragments of a single protein which when united result in a simple biological readout – the two protein frag- ments are fused to the potential bait/prey interacting partners respectively. This methodology is easily scalable to HT since it can 2.3. Proteomics High-Throughput Technologies 2.3.1. Mass Spectrometry High-Throughput Technologies 2.3.2. Interactomics HT Technologies
- 26. 7 Omics Technologies, Dataand Bioinformatics Principles yield very high numbers of coding sequences assayed in a relatively simple experiment and a wide variety of interactions can be detected and characterised following one single, commonly used protocol. However, the proteins are being expressed in an alien cell system with a loss of temporal and physiological control of expression patterns, resulting in a large number of false-positive interactions. Affinity-based assays, such as affinity chromatogra- phy, pull-down and coimmunoprecipitation, rely on the strength of the interaction between two entities. These techniques can be used on interactions which form under physiological conditions, but are only as good as the reagents and techniques used to iden- tify the participating proteins. High throughput mass spectrom- etry is increasingly used for the rapid identification of the participants in an affinity complex. Physical methods depend on the properties of molecules to enable measurement of an interac- tion, as typified by techniques such as X-ray crystallography and enzymatic assays. High quality data can be produced but highly purified proteins are required, which has always proved a rate limiting step. Availability of automated chromatography systems and custom robotic systems that streamline the whole process, from cell harvesting and lysis through to sample clarification and chromatography has changed this, and increasing amounts of data are being generated by such experiments. It is now largely the case that high throughput methods exist for all or most of the Omics domains. The challenge now is to prevent bottlenecks appearing in the storing, annotation, and analysis of the data. First the data which is required to describe both – how an experiment was performed and the results generated by it – must be defined. A place to store that information must be identi- fied, a means by which it will be gathered has to be agreed upon, and ways in which the information will be queried, retrieved and analysed must also be decided. Data in isolation is of limited use, so ideally the data format chosen should be appropriate to enable the combination and comparison of multiple datasets, both in-house and with other groups working in the same area. HT data is increasingly used in a broader context beyond the indi- vidual project; consequently it is becoming more important to standardise and share this information appropriately and to pre- interpret it for the scientists who are not involved with the experi- ment, whilst still making the raw data available for those who wish to perform their own analyses. In high throughput research, knowledge discovery starts by collecting, selecting and cleaning the data in order to fill a data- base. A database is a collection of files (archive) of consistent data that are stored in a uniform and efficient manner. A relational database consists of a set of tables, each storing records (instances). 2.4. Challenges in HT Technologies 2.5. Bioinformatics Concepts
- 27. 8 Schneider andOrchard A record is represented as a set of attributes which define a property of a record. Attributes can be identified by their name and store a value. All records in a table have the same number and type of attributes. Database design is a crucial step in which the data requirements of the application have first to be defined (concep- tual design), including the entities and their relationships. Logical design is the implementation of the database using database management systems, which ensure that the process is scalable. Finally the physical design phase estimates the workload and refines the database design accordingly. It is during this phase that table designs are optimized, indexing is implemented and cluster- ing approaches are optimized. These are fundamental in order to obtain fast responses to frequent queries without jeopardising the database integrity (e.g. redundancy). Primary or archived data- bases contain information directly deposited by submitters and give an exact representation of their published data, for example DNA sequences, DNA and protein structures and DNA and pro- tein expression profiles. Secondary or derived databases are so- called because they contain the results of analysis on the primary resources, including information on sequence patterns or motifs, variants and mutations and evolutionary relationships. The fundamental characteristic of a database record is a unique identifier. This is crucial in biology given the large num- ber of situations where a single entity has many names, or one name refers to multiple entities. To some extent, this problem can be overcome by the use of an accession number, a primary key derived by a reference database to describe the appearance of that entity in that database. For example, using the UniProtKB pro- tein sequence database accession number of human p53 gene products (P04637) gives information on the sequence of all the isoforms of these proteins, gene and protein nomenclature as well as a wealth of information about its function and role in a cell. More than one protein sequence database exists, and the vast majority of protein sequences exist in all of these. Fortunately resources to translate between these multiple accession numbers now exist, for example the Protein Identifier Cross-Reference (PICR) Service at the European Bioinformatics Institute (EBI) (see Note 8). The Omics fields share with all of biology the challenge of handling ever-increasing amounts of complex information effec- tively and flexibly. Therefore a crucial step in bioinformatics is to choose the appropriate representation of the data. One of the simplest but most efficient approaches has been the use of controlled vocabularies (CVs), which provide a standardised dictionary of terms for representing and managing information. Ontologies are structure CVs. An excellent example of this methodology is the Gene Ontology (GO) that describes gene products in terms of their associated biological processes, cellular
- 28. 9 Omics Technologies, Dataand Bioinformatics Principles components and molecular functions in a species independent manner. Substantial effort has been, and continues to be, put into the development and maintenance of the ontologies themselves; the annotation of gene products, which entails making associa- tions between the ontologies and the genes and gene products across databases; and the development of tools that facilitate the creation, maintenance and use of ontologies. The hierarchical nature of these CVs enable more meaningful queries to be made, for example searching either a microarray or proteomics experi- ment for expression patterns on the brain, enable experiments annotated to the cortex to be included because the BRENDA tis- sue CV recognises the cortex as “part-of” the brain (http:/ /www. ebi.ac.uk/ontology-lookup/browse.do?ontName=BTO). Use of these CVs have been encouraged, and even made mandatory, by many groups such as the Microarray Expression Data group (MGED) which recommends the use of the MGED ontology (6) for the description of key experimental concepts and, where pos- sible, ontologies developed by other communities for describing terms such as anatomy, disease and chemical compounds. Clustering methods are used to identify patterns in the data, in other words to recognise what is similar, to identify what is dif- ferent, and from there to know when differences are meaningful. These three steps are not trivial at all; proteins for example exhibit rich evolutionary relationships and complex molecular interac- tions and hence present many challenges for computational sequence analysis. Sequence similarity refers to the degree to which nucleotide or protein sequences are related. The extent of similarity between two sequences can be based on percent sequence identity (the extent to which two (nucleotide or amino acid) sequences are invariant) and/or conservation (changes at a specific position of an amino acid or nucleotide sequence that preserve the physicochemical properties of the original residue). The applications of sequence similarity searching are numerous, ranging from the characterization of newly sequenced genomes, through phylogenetics, to species identification in environmental samples. However, it is important to keep in mind that identify- ing similarity between sequences (e.g. nucleotide or amino acid sequences) is not necessarily equivalent to identifying other pro perties of such sequences, for example their function. It is obvious that without bioinformatics it is impossible to make sense of the huge data produced in Omics research. If we look at the increase of the EMBL Nucleotide Sequence Database (EMBL-Bank), the Release 105 on 27-AUG-2010 contained 3. Methods
- 29. 10 Schneider andOrchard 195,241,608 sequence entries comprising 292,078,866,691 nucleotides. This translated to a total of 128 GB compressed and 831 GB uncompressed data. Bioinformatics does not only have to provide the structures in which to store the information, but also store it in such a way that is retrievable, and comparable not only to similar data but also to other types of information. The challenges and concepts bioinformatics as a discipline currently encompasses do not essentially differ from those listed by (7), they have merely expanded to meet the challenges imposed by the volume of data produced. These include: 1. A precise, predictive model of transcription initiation and ter- mination: the ability to predict where and when transcription will occur in a genome (fundamental for HTS and proteomics); 2. A precise, predictive model of RNA splicing/alternative splic- ing: the ability to predict the splicing pattern of any primary transcript in any tissue (fundamental for transcriptomics and proteomics); 3. Precise, quantitative models of signal transduction pathways: ability to predict cellular responses to external stimuli (required in proteomics and pathways analysis); 4. Determination of effective protein:DNA, protein:RNA and protein:protein recognition codes (important for recognition of interactions among the various types of molecules); 5. Accurate ab initio protein structure prediction (required for proteomics and pathways analysis); 6. Rational design of small molecule inhibitors of proteins (chemogenomics); 7. Mechanistic understanding of protein evolution: understand- ing exactly how new protein functions evolve (comparative genomics); 8. Mechanistic understanding of speciation: molecular details of how speciation occurs (comparative genome sequences, sequence variation); 9. Continued development of effective gene ontologies – sys- tematic ways to describe the functions of any gene or protein (genomics, transcriptomics, and proteomics). The above list summarises general concepts required for mul- tiple Omics data sources. Next we describe issues which are spe- cifictooneparticularfield,butmayhavedownstreamconsequences in other areas. Here we will explore two major challenges in genomics: de novo sequencing assembly and genome annotation. 3.1. The Role of Bioinformatics in Genomics
- 30. 11 Omics Technologies, Dataand Bioinformatics Principles A critical stage in de novo genome sequencing is the assembly of shotgun reads, in other words putting together fragments randomly extracted from the sample to form a set of contiguous sequences and contigs that represent the DNA in the sample. Algorithms are available for whole genome shotgun fragment assembly, including Atlas (8), Arachne (9), Celera (10), PCAP (11), Phrap (http:/ /www.phrap.org) and Phusion (12). All these programmes rely on the overlap-layout-consensus approach (13) where all the reads are compared to each other in a pair-wise fash- ion. However, this approach presents several disadvantages, espe- cially in the case of next-generation microread sequencing. EDENA (14) is the only microread assembler developed using computation of pairwise overlaps. Included reads, i.e. reads which align over their whole length onto another read, have to be removed from the graph; this means that mixed-length sequenc- ing cannot be performed directly with an overlap graph. Short reads are either simply mapped onto long read contigs or they are assembled separately (Daniel Zerbino personal communication). The use of a sequence graph to represent an assembly was introduced by (15). Idury and Waterman presented an assembly algorithm for an alternative sequencing technique, sequencing by hybridisation, where an oligoarray could detect all the k nucle- otide words, also known as k-mers, present in a given genome. Pevzner et al. (16) expanded on this idea, proposing a slightly different formalisation of the sequence graph, called the de Bruijn graph, whereby the k-mers are represented as arcs and overlap- ping k-mers join at their tips, and consecutively presented algo- rithms to build and correct errors in the de Bruijn graph (13), use paired-end reads (16) or short reads (17). Zerbino and Birney (18) developed a new set of algorithms, collectively called “Velvet,” to manipulate de Bruijn graphs for genomic sequence assembly for the de novo assembly of microreads. Several studies have used Velvet (19–22). Other analytical software adopting the use of the de Bruijn graph are ALLPATHS (23) and SHORTY (24) specialised in localising the use of paired-end reads, whereas the ABySS (25, 26) successfully parallelised the construction of the de Bruijn graph, thus removing practical memory limitations on assemblies. The field of de novo assembly of NGS reads is constantly evolving and there is not yet a firm process or best practise set in place. Genome annotation is the process of marking the genes and other biological features in a DNA sequence. It consists of two main steps: (1) Gene Finding: identifying elements on the genome and (2) adding biological information to these elements. There are automatic annotation tools to perform all this by computer analy- sis, as opposed to manual annotation which involves human expertise. Ideally, these approaches coexist and complement each 3.1.1. De Novo Genome Sequencing 3.1.2. Genome Annotation
- 31. 12 Schneider andOrchard other in the same annotation pipeline. The basic level of annotation uses BLAST to find similarities, and annotates genomes based on that. However, nowadays more and more additional information is added to the annotation platform. Structural annotation con- sists of the identification of genomic elements: ORFs and their localisation, gene structure, coding regions and location of regu- latory motifs. Functional annotation consists in attaching biologi- cal information to genomic elements: biochemical function, biological function, involved regulation and interactions and expression. These steps may involve both biological experiments and in silico analysis and are often initially performed in related databases, usually protein sequence databases such as UniProtKB, and transferred back onto the genomic sequence. A variety of software tools have been developed to permit scientists to view and share genome annotations. The additional information allows manual annotators to disentangle discrepan- cies between genes that have been given conflicting annotation. For example, the Ensembl genome browser relies on both curated data sources as well as a range of different software tools in their automated genome annotation pipeline (27). Genome annota- tion remains a major challenge for many genome projects. The identification of the location of genes and other genetic control elements is frequently described as defining the biological “parts list” for the assembly and normal operation of an organism. Researchers are still at an early stage in the process of delineating this parts list, as well as trying to understand how all the parts “fit together”. Both microarray and proteomics experiments provide long lists of transcripts (mRNA and proteins respectively) co-expressed at any one time and the challenge is to give biological relevance to these lists. Several different computational algorithms have been devel- oped and can be usefully applied at various steps of the analytical pipeline. Clustering methods are used to order and visualise the underlying patterns in large scale expression datasets showing similar patterns that can therefore be grouped according to their co-regulation/co-expression (e.g. specific developmental times or cellular/tissue locations). This indicates (1) co-regulated tran- scripts which might be functionally related and (2) the clusters represent a natural structure of the data. Transcripts can also be grouped by their known – or predicted function. A resource commonly used for this is the GO ontology (http://www.geneontology.org). There are several bioinformat- ics tools for calculating the number of significantly enriched GO terms, for example: (1) GO miner (http://discover.nci.nih.gov/ gominer) generates a summary of GO terms that are significantly enriched in a user input list of protein accession numbers when compared to a reference database like UniProtKB/SwissProt; 3.2. The Role of Bioinformatics in Transcriptomics
- 32. 13 Omics Technologies, Dataand Bioinformatics Principles (2) GO slims which are subsets of GO terms from the whole Gene Ontology and are particularly useful for giving a summary of the results of GO annotation of a genome, microarrays and proteomics (http:/ /amigo.geneontology.org/cgi-bin/amigo/ go.cgi). The use of different bioinformatics approaches to determine the presence of a gene or open reading frame (ORF) in those genomes can lead to divergent ORF annotations (even for data generated from the same genomic sequences). It is therefore crucial to use the correct dataset for protein sequence translations. One method for confirming a correct protein sequence is mass spectrometry based proteomics, in particular by de novo sequencing which does not rely on pre-existing knowledge of a protein sequence. However, historically, there has initially been no method for pub- lishing these protein sequences, except as long lists reported directly with the article or included on the publisher’s website as supplementary information. In either case, these lists are typically provided as PDF or spreadsheet documents with a custom-made layout, making it practically impossible for computer programmes to interpret them, or efficiently query them. A solution to this problem is provided by the PRIDE database (http:/ /www.ebi.ac. uk/pride) which provides a standards compliant, public reposi- tory for mass spectrometry based proteomics, giving access to experimental evidence that a transcribed gene product does exist, as well as the pattern of tissues in which it is expressed (28). The annotation of protein functional information largely relies on manual curation, namely biologists reading the scientific literature and transferring the information to computational records – a process in which the UniProtKB curators have lead the way for many years. The many proteins for which functional information is not available, however, rely on selected informa- tion being transferred from closely related orthologues in other species. A number of protein signature databases now exist, which create algorithms to recognise these closely related protein fami- lies or domains within proteins. These resources have been com- bined in a single database, Interpro (http:/ /www.ebi.ac.uk/ interpro) and the tool InterProScan (see Note 9) (http:/ /www. ebi.ac.uk/Tools/InterProScan) is available for any biologist wishing to perform their own automated protein (or gene) anno- tation (29). Protein–protein interactions are generally represented in graphi- cal networks with nodes corresponding to the proteins and edges to the interactions. Although edges can vary in length most net- works represent undirected and only binary interactions. Bioinformatics tools and computational biology efforts into graph theory methods have and continue to be part of the knowledge 3.3. The Role of Bioinformatics in Proteomics 3.3.1. Protein Annotation 3.3.2. Protein–Protein Interaction Analysis and Comparative Interactomics
- 33. 14 Schneider andOrchard discovery process in this field. Analysis of PPI networks involves many challenges, due to the inherent complexity of these net- works, high noise level characteristic of the data, and the presence of unusual topological phenomena. A variety of data-mining and statistical techniques have been applied to effectively analyze PPI data and the resulting PPI networks. The major challenges for computational analysis of PPI networks remain: 1. Unreliability of large scale experiments; 2. Biological redundancy and multiplicity: a protein can have several different functions; or a protein may be included in one or more functional groups. In such instances overlapping clusters should be identified in the PPI networks, however since conventional clustering methods generally produce pairwise disjoint clusters, they may not be effective when applied to PPI networks; 3. Two proteins with different functions frequently interact with each other. Such frequent connections between the proteins in different functional groups expand the topological com- plexity of the PPI networks, posing difficulties to the detec- tion of unambiguous partitions. Intensive research trying to understand and characterise the structural behaviours of such systems from a topological perspec- tive have shown that features such as small-world properties (any two nodes can be connected via a short path of a few links), scale- free degree distributions (power-law degree distribution indicat- ing that a few hubs bind numerous small nodes), and hierarchical modularity (hierarchical organization of modules) suggests that a functional module in a PPI network represents a maximal set of functionally associated proteins. In other words, it is composed of those proteins that are mutually involved in a given biological process or function. In this model, the significance of a few hub nodes is emphasized, and these nodes are viewed as the determi- nants of survival during network perturbations and as the essen- tial backbone of the hierarchical structure. The information retrieved from HT interactomics data could be very valuable as a means to obtain insights into a systems evolu- tion (e.g. by comparing the organization of interaction networks and by analyzing their variation and conservation). Likewise, one could learn whether and how to extend the network information obtained experimentally in well-characterised model systems onto different organisms. Cesareni et al. (30) concluded that, despite the recent completion of several high throughput experiments aimed at the description of complete interactomes, the available interaction information is not yet of sufficient coverage and qual- ity to draw any biologically meaningful conclusion from the com- parison of different interactomes. The development of more
- 34. 15 Omics Technologies, Dataand Bioinformatics Principles accurate experimental and informatics approaches is required to allow us to study network evolution. The massive amounts of data produced in Omics experiments can help us gain insights into underlying biological processes only if they are carefully recorded and stored in databases, where they can be queried, compared and analyzed. Data has to be stored in a structured and standardized format that enables data sharing between multiple resources, as well as common tool development and the ability to merge data sets generated by different tech- nologies. Omics is very much technology driven, and all instru- ment and software manufacturers initially produce data in their own proprietary formats, often then tying customers into a lim- ited number of downstream analytical instruments. Efforts have been ongoing for many years to develop and encourage the devel- opment of common formats to enable data exchange and stan- dardized methods for the annotation of such data to allow dataset comparison. These efforts were spear-headed by the transcriptomics community, who developed the MIAME standards (Minimum Information About a Microarray Experiment, http:/ /www.mged. org/Workgroups/MIAME/miame.html) (31). The MIAME standards describe the set of information sufficient to interpret a microarray experiment and its results unambiguously, to enable verification of the data and potentially to reproduce the experi- ment itself. Their lead was soon followed by the proteomics community with the MIAPE standards (Minimum Information About a Proteomics Experiment, http:/ /www.psidev.info/index. php?q=node/91), the interaction community (MIMIx, http:/ / imex.sourceforge.net/MIMIx) and many others. This has resulted in the development of tools which can combine datasets, for example it is possible to import protein interaction data into the visualisation tool Cytoscape (http:/ /www.cytoscape.org) in a common XML format (PSI-MI) and overlay this with expression data from a microarray experiment. Below we will follow the three Omics fields we described above. It would be impossible to list all the databases dealing with these data, however as the European Bioinformatics Institute hosts one of the most comprehensive sets of bioinformatics databases and also actively coordinates or is involved in setting standards and their implementation, it serves as exemplar for databases that are at the state of the art for standards, technologies and integration of the data. A list of major Institutes and their databases is pro- vided at the end of this chapter (see Note 18). The genome is a central concept at the heart of biology. Since the first complete genome was sequenced in the mid-1990s, over 800 3.4. Storing Omics Data Appropriately 3.5. Exploring Omics Data in Bioinformatics 3.5.1. Genomics
- 35. 16 Schneider andOrchard more have been sequenced, annotated, and submitted to the public databases. New ultra-high throughput sequencing tech- nologies are now beginning to generate complete genome sequence at an accelerating rate, both to gap-fill portions of the taxonomy where no genome sequence has yet been deciphered (e.g. the GEBA project, http:/ /www.jgi.doe.gov/programs/ GEBA, which aims to sequence 6,000 bacteria from taxonomi- cally distinct clades), and to generate data for variation in popula- tions of species of special interest (e.g. the 1000 Genomes Project inhuman,http:/ /www.1000genomes.org,andthe1001Genomes Project in Arabidopsis, http:/ /www.1001genomes.org). In addi- tion, modern sequencing technologies are increasingly being used to generate data for gene regulation and expression on a genome- wide scale. The vast amount of information associated with the genomic sequence demands a way to organise and access it (see Note 19). A successful example of this is the genome browser Ensembl. Ensembl (http:/ /www.ensembl.org) is a joint project between the EBI and the Wellcome Trust Sanger Institute that annotates chordate genomes (i.e. vertebrates and closely related invertebrates with a notochord such as sea squirt). Gene sets from model organisms such as yeast and fly are also imported for comparative analysis by the Ensembl “compara” team. Most annotation is updated every 2 months; however, the gene sets are determined about once a year. A new browser, http:/ /www. ensemblgenomes.org, has now been set up to access non- chordates genomes from bacteria, plants, fungi, metazoa and protists. Ensembl provides genes and other annotation such as regula- tory regions, conserved base pairs across species, and mRNA pro- tein mappings to the genome. Ensembl displays many layers of genome annotation into a simplified view for the ease of the user. The Ensembl gene set reflects a comprehensive transcript set based on protein and mRNA evidence in UniProt and NCBI RefSeq databases (see Note 10). These proteins and mRNAs are aligned against a genomic sequence assembly imported from a relevant sequencing centre or consortium. Transcripts are clus- tered into the same gene if they have overlapping coding sequence. Each transcript is given a list of mRNAs and proteins it is based upon. Ensembl utilises BioMart, a query optimised database for effi- cient data mining described below, and the application of a com- parative analysis pipeline: Compara. The Ensembl Compara multi-species database stores the results of genome-wide species comparisons calculated for each data release including: (1) Comparative genomics: Whole genome alignments and Synteny regions and (2) Comparative proteomics: Orthologue predictions and Paralogue predictions.
- 36. 17 Omics Technologies, Dataand Bioinformatics Principles Ensembl Compara includes GeneTrees, a comprehensive gene orientated phylogenetic resource. It is based on a computa- tional pipeline to handle clustering, multiple alignment, and tree generation, including the handling of large gene families. Ensembl also imports variations including Single Nucleotide Polymorphisms and insertion-deletion mutations (Indels) and their flanking sequence from various sources. These sequences are aligned to the reference sequence. Variation positions are calcu- lated in this way along with any effects on transcripts in the area. The majority of variations are obtained from NCBI dbSNP. For human, other sources include Affymetrix GeneChip Arrays, The European Genome-phenome Archive, and whole genome align- ments of individual sequences from Venter (32), Watson (33) and Celera individuals (34). Sources for other species include Sanger re-sequencing projects for mouse, and alignments of sequences from the STAR consortium for rat. Ancestral alleles from dbSNP were determined through a comparison study of human and chimpanzee DNA (35). There is a wide range of HT transcriptomics data: single and dual channel microarray-based experiments measuring mRNA, miRNA and generally non-coding RNA. One can also include non-array techniques such as serial analysis of gene expression (SAGE). There are three main public repositories on microarray based studies: ArrayExpress (36), Gene Expression Omnibus (37), and CIBEX (38). Here we describe the EBI microarray repository, ArrayExpress, which consists of three components: the ArrayExpress Repository – a public archive of functional ● ● genomics experiments and supporting data, the ArrayExpress Warehouse – a database of gene expression ● ● profiles and other bio-measurements, the ArrayExpress Atlas – a new summary database and meta- ● ● analytical tool of ranked gene expression across multiple experiments and different biological conditions. The Warehouse and Atlas allow users to query for differen- tially expressed genes by gene names and properties, experimental conditions and sample properties, or a combination of both (39). The latest developed ArrayExpress Atlas of Gene Expression (http:/ /www.ebi.ac.uk/microarray-as/atlas) allows the user to query for condition-specific gene expression across multiple data sets. The user can query for a gene or a set of genes by name, synonym, Ensembl identifier, GO term or, alternatively, for a bio- logical sample property or condition, (e.g. tissue type, disease name, developmental stage, compound name or identifier). Queries for both genes and conditions are also possible (e.g. the user can query for all “DNA repair” genes up-regulated in 3.5.2. Transcriptomics
- 37. 18 Schneider andOrchard “cancer” which returns a list of “experiment, condition, gene” triplets each with a P-value and an up/down arrow characterising the significance and direction of a gene’s differential expression in a particular condition in an experiment). ArrayExpress accepts data generated on all array-based tech- nologies, including gene expression, protein array, ChIP-chip and genotyping. More recently, data from transcriptomic and related applications of uHTS technologies such as Illumina (SOLEXA Ltd, Saffron Walden, UK), and 454 Life Sciences (Roche, Branford, Connecticut) are also accepted. For Solexa data FASTQ files, sample annotation and processed data files corresponding to transcription values per genomic location are submitted and curated to the emerging standard MINSEQE (http:/ /www.mged. org/minseqe) and instrument-level data are stored in the European Short Read Archive (http:/ /www.ebi.ac.uk/embl/ Documentation/ENA-Reads.html).TheArrayExpressWarehouse now includes gene expression profiles from in situ gene expres- sion measurements, as well as other molecular measurement data from metabolomics and protein profiling technologies. Where in situ and array-based gene expression data are available for the same gene, these are displayed in the same view and links are pro- vided to the multispecies 4DXpress database of in situ gene expression (39). The Gene Expression Atlas provides a statistically robust framework for integration of gene expression experiment results across different platforms at a meta-analytical level. It also repre- sents a simple interface for identifying strong differential expres- sion candidate genes in conditions of interest. The Atlas also integrates ontologies for high quality annotation of gene and sample attributes and builds new gene expression summarised views, with the aim to provide analysis of putative signalling path- way targets, discovery of correlated gene expression patterns and the identification of condition/tissue-specific patterns of gene expression. A list of URLs to bioinformatics relevant resources to transcriptomics can be found in Subheading 4 (see Note 20). A list of proteomics relevant bioinformatics resources can be found in Note 21. Translated proteins and their co-translational modification or PTM (post-translated modifications) are the backbone of pro- teomics (28). UniProt is the most comprehensive data repository on protein sequence and functional annotation. It is maintained by a collaboration of the Swiss Institute of Bioinformatics (SIB), the Protein Information Resource (PIR), and the EBI. It has four components, each of them optimized for different user profiles: 1. UniProt Knowledgebase (UniProtKB) comprises two sec- tions: UniProtKB/Swiss-Prot and UniProtKB/TrEMBL. 3.5.3. Proteomics 3.5.3.1. Protein Sequence and Functional Annotation
- 38. 19 Omics Technologies, Dataand Bioinformatics Principles (a) UniProtKB/Swiss-Prot contains high quality annotation extracted from the literature and computational analyses curated by experts. Annotations include, amongst oth- ers: protein function(s), protein domains and sites, PTMs, subcellular location(s), tissue specificity, structure, inter- actions, and diseases associated with deficiencies or abnormalities. (b) UniProtKB/TrEMBL contains the translations of all coding sequences (CDS) present in the EMBL/ GenBank/DDJB nucleotide sequence databases, exclud- ing some types of data such as pseudogenes. UniProtKB/ TrEMBL records are annotated automatically based on computational analyses. 2. UniProt Reference Clusters (UniRef), which provides clus- tered sets of all sequences from the UniProtKB database and selected UniProt Archive records to obtain complete cover- age of sequences at different resolutions (100, 90, and 50% sequence identity), while hiding redundant sequences. 3. UniProt archive (UniParc) is a repository that reflects the his- tory of all protein sequences. 4. UniProt Metagenomic and Environmental Sequences data- base (UniMES) contains data from metagenomic projects such as the Global Ocean Sampling Expeditions. UniProtKB includes cross-references from over 120 external databases, including Gene Ontology (GO), InterPro (protein families and domains), PRIDE (Protein identification data), IntEnz (enzyme) (see Note 11), OMIM (the Online Mendelian Inheritance in Man database) (see Note 12), Interaction data- bases (e.g. IntAct, DIP, Mint, see Note 21), Ensembl, several genomic databases from potential pathogens (e.g. EchoBase, Ecogene, LegioList, see Note 13), the European Hepatitis C Virus database (http:/ /euhcvdb.ibcp.fr/euHCVdb) and others. Several repositories have been established to store protein and peptide identifications derived from MS, the main method for the identification and quantification of proteins (28). There are two main repositories for MS data in proteomics: The Proteomics IDEntifications database (PRIDE, ● ● http:/ / www.ebi.ac.uk/pride) Peptidome ( ● ● http:/ /www.ncbi.nlm.nih.gov/peptidome)and a number of related resources such as PeptideAtlas (http:/ / www.peptideatlas.org) and the Global Proteomics Machine (http:/ /www.thegpm.org/GPMDB), which take deposited raw data for reanalysis in their own pipeline. These all serve as web-based portals for data mining, data visu- alisation, data sharing, and cross-validation resources in the field. 3.5.3.2. Mass Spectrometry Repositories
- 39. 20 Schneider andOrchard The proteomics identifications (PRIDE) database has been built to turn publicly available data, buried in numerous academic pub- lications, into publicly accessible data. PRIDE is fully compliant to the standards released by the HUPO-PSI and also makes extensive use of CVs such as Taxonomy, the BRENDA Tissue Ontology and Gene Ontology, thus direct access to PRIDE data organised by species, tissue, sub-cellular location, disease state and project name can be obtained via the “Browse Experiments” menu item. PRIDE remains the most complete database in terms of metadata associated with peptide identifications, since it con- tains numerous experimental details of the protocols followed by the submitters (28). The detailed metadata in PRIDE has enabled analyses of large datasets which have proven to yield very interest- ing information for the field (28). PRIDE uses Tranche (see Note 14) to allow the sharing of massive data files, currently including search engine output files and binary raw data from mass spectrometers that can be accessed via a hyperlink from PRIDE. As a member of the Proteome Exchange consortium PRIDE will make both the annotated meta-raw spectral data available, via Tranche to related analytical pipelines such as PeptideAtlas (see Note 15) and The Global Proteome Machine (see Note 16). Both the number of laboratories producing PPI data and the size of such experiments continues to increase and a number of repos- itories exist to collect this data (see Note 22). Here we explore IntAct, a freely available, open source database system and analy- sis tools for molecular interaction data derived from literature or direct user submissions. IntAct follows a deep curation model, capturing a high level of detail from the experimental reports on the full text of the publication. Queries may be performed on the website with the initial presentation of the data as a list of binary interaction evidences. Users can access the individual evidences that describe the interaction of two specific molecules, thus allow- ing users to filter result sets (e.g. by interaction detection method) to only retain user-defined evidences. For convenience, evidence pertaining to the same interactors is grouped together in the binary interaction evidence table. Downloads of any datasets are available in both PSI-MI XML and tab-delineated MITAB for- mat, providing end users with the highest level of details without compromising the integrity and simplicity of access to the data (40). IntAct is also involved in a major data exchange collabora- tion driven by the major public interaction data providers (listed at the end of this Chapter): The International Molecular Exchange Consortium (IMEx, http:/ /imex.sourceforge.net) partners share curation efforts and exchange completed records on molecular interaction data. Curation has been aligned to a common standard, as detailed in the curation manual of the individual databases and 3.5.3.3. Protein–Protein Interactions and Interactomics
- 40. 21 Omics Technologies, Dataand Bioinformatics Principles summarised in the joint curation manual available at http:/ /imex. sourceforge.net. IMEx partner databases request the set of mini- mum information about a molecular interaction experiment (MIMIx) to be provided with each data deposition (41). The use of common data standards encourages the develop- ment of tools utilising this format. For example Cytoscape (http:/ / www.cytoscape.org) resembles an open source bioinformatics software platform for visualising molecular interaction networks and integrating these interactions with gene expression profiles and other state data, in which data from resources such as IntAct can be visualised and combined with other datasets. The value of the information obtained from comparing net- works depends heavily on both the quality of the data used to assemble the networks themselves and the coverage of these net- works (30, 42). The most comprehensive studies are in Saccharomyces cerevisiae; however, it should be noted that two comparable, “comprehensive” experiments, performed in parallel by two different groups using the same approach (tandem affinity purification technology) ended up with fewer than 30% of the interactions discovered by each group in common (43), suggest- ing that coverage is far from complete. In the Omics several efforts have been and continue to be made in order to create computational tools for integrating Omics data. These need to address three different aspects of the integration (44): 1. To identify the network scaffold by delineating the connec- tions that exist between cellular components; 2. To decompose the network scaffold into its constituent parts in an attempt to understand the overall network structure; 3. To develop cellular or system models to simulate and predict the network behaviour that gives rise to particular cellular phenotypes. As we have seen in the previous section here are significant challenges to modern post-genomics data sets: 1. Many technological platforms, both hardware and software, are available for several Omics data types, but some of these are prone to introducing technical artefacts; 2. Standardized data representations are not always adopted, which complicates cross-experiment comparisons; 3. Data-quality, context and lab-to-lab variations represent another important hurdle that must be overcome in genome- scale science. Obviously the spread of Omics data in wide variety of formats represents a challenge for encompassing the technical hitches in integrating and migrating across platforms. One of the important 3.6. Integration of Omics Data
- 41. 22 Schneider andOrchard techniques often used is XML. XML is used to provide a document markup language that is easier to learn, retrieve, store and trans- mit. It is semantically richer than HTML (45). Here we present three different infrastructures which have been used and repre- sent different ways of integration of Omics data: BioMart, Taverna and the BII Infrastructure. BioMart is a query-oriented DBMS developed jointly by the Ontario Institute for Cancer Research and the EBI: BioMart (http:/ /www.biomart.org) is particularly suited for providing “data mining” like searches of complex descriptive data. It can be used with any type of data as shown by some of the resources cur- rently powered by BioMart: Ensembl, UniProt, InterPro, HGNC, RatGenomeDatabase,ArrayExpressDW,HapMap,GermOnLine, PRIDE, PepSeeker, VectorBase, HTGT and Reactome. BioMart comes with an “out of the box” website that can be installed, configured and customised according to user require- ments. Further access is provided by graphical and text based applications or programmatically using web services or API writ- ten in Perl and Java. BioMart has built-in support for query opti- misation and data federation and in can also be configured to work as a DAS 1.5 Annotation server. The process of converting a data source into BioMart format is fully automated by the tools included in the package. Currently supported RDBMS platforms are MySQL, Oracle and Postgres. BioMart is completely Open Source, licenced under the LGPL, and freely available to anyone without restrictions (46). The Taverna workbench (http:/ /taverna.sourceforge.net) is a free software tool for designing and executing workflows, cre- ated by the myGrid project (http:/ /www.mygrid.org.uk/tools/ taverna), and funded through OMII-UK (http:/ /www.omii. ac.uk). Taverna allows users to integrate many different software tools, including web services from many different domains. Bioinformatics services include those provided by the National Centre for Biotechnology Information, The EBI, the DNA Databank of Japan, SoapLab, BioMOBY and EMBOSS (see Note 17). Effectively, Taverna allows a scientist with limited computa- tional background and technical resource support to construct highly complex analyses over public and private data and compu- tational resources, all from a standard PC, UNIX box or Apple computer. A successful example of using Taverna in Omics is demonstrated by the work of Li et al. (47) where the authors describe an example of a workflow involving the statistical identi- fication of differentially expressed genes from microarray data fol- lowed by the annotation of their relationships to cellular processes. They show that Taverna can be used by data analysis experts as a 3.6.1. BioMart 3.6.2. Taverna
- 42. 23 Omics Technologies, Dataand Bioinformatics Principles generic tool for composing ad hoc analyses of quantitative data by combining the use of scripts written in the R programming language with tools exposed as services in workflows (47). As we have seen, it is now possible to run complex multi-assay studies through a variety of Omics technologies, for example determining the effect on a number of subjects, of a compound by characterising a metabolic profile (by mass spectroscopy), mea- suring tissue specific protein and gene expression (by mass spec- trometry and DNA microarrays, respectively), and conducting conventional histological analysis. It is essential that such com- plex metadata (i.e. sample characteristics, study design, assay exe- cution, sample-data relationships) are reported in a standard manner to correctly interpret the final results (data) that they contextualise. Relevant EBI systems, such as ArrayExpress, PRIDE and ENA-Reads (The European Nucleotide Archive (ENA) accepts data generated by NGS methodologies such as 454, Illumina and ABI SOLiD) are built to store microarray- based, proteomics and NGS-based experiments, respectively. However, these systems have different submission and download formats, and diverse representations of the metadata and termi- nologies used. Nothing yet exists to archive metabolomics-based assays and other conventional biomedical/environmental assays. The BioInv Index (BioInvestigation Index, http:/ /www.ebi. ac.uk/net-project/projects.html) infrastructure (BII) aims to fill this gap. BII infrastructure aims to be a single entry point for those researchers willing to deposit their multi-assay studies and datasets, and/or easily download similar datasets. This infrastruc- ture allows commonly representing and storing the experimental metadata of biological, biomedical and environmental studies. Although relying on other EBI production systems, the BII infra- structure shields the users from their diverse formats and ontologies, by progressively implementing in the editor tool integrative cross- domain “standards” such as MIBBI, OBO Foundry and ISA-TAB. A prototype instance is up and running at http:/ /www.ebi.ac.uk/ bioinvindex/home.seam. 1. Wellcome Trust Sanger Sequencing Centre: The Sanger Institute is a genome research institute primarily funded by the Wellcome Trust. The Sanger uses large-scale sequencing, informatics and analysis of genetic variation to further improve our understanding of gene function in health and disease and to generate data and resources of lasting value to biomedical research, see http:/ /www.sanger.ac.uk. 3.6.3. BII Infrastructure 4. Notes
- 43. 24 Schneider andOrchard 2. Metagenomics: The term indicates the study of metagenomes, genetic material recovered directly from environmental sam- ples. It is also used generically for environmental genomics, ecogenomics or community genomics. Metagenomics data can be submitted and stored in appropriate databases (see http:/ /www.ncbi.nlm.nih.gov/Genbank/metagenome.html and http:/ /www.ebi.ac.uk/genomes/wgs.html). 3. Metatranscriptomics: This term refers to studies where micro- bial gene expression in the environment is accessed (e.g. pyrosequencing) directly from natural microbial assemblages. 4. Epigenomics: Understanding the large numbers of variations in DNA methylation and chromatin modification by exploit- ing omics techniques. There are various recent efforts in this direction (i.e. http:/ /www.heroic-ip.eu). 5. Studies of genome variation: Clear examples on the advances on this front come from the large-scale human variation data- bases which archive and provide access to experimental data resulting from HT genotyping and sequencing technologies. The European Genotype Archive (http:/ /www.ebi.ac.uk/ ega/page.php) provides dense genotype data associated with distinct individuals. Another relevant projects on this front is ENCODE (http:/ /www.genome.gov/10005107), the Encyclopedia Of DNA Elements, which aims to identify all functional elements in the human genome sequence. 6. Cycle-array sequencing methods: also known as NGS: Cycle- array methods generally involve multiple cycles of some enzymatic manipulation of an array of spatially separated oligonucleotide features. Each cycle only queries one or a few bases, but an enormous number of features are processed in parallel. Array features can be ordered or randomly dispersed. 7. Next generation expressed-sequence-tag sequencing: ESTs are small pieces of DNA sequence (200–500 nucleotides long) that are generated by sequencing of an expressed gene. Bits of DNA that represent genes expressed in certain cells, tissues, or organs from different organisms are sequenced and use as “tags” to fish a gene out of a portion of chromosomal DNA by matching base pairs. Characterising transcripts through sequences rather than hybridization to a chip has its advantages (i.e. the sequencing approach does not require the knowledge of the genome sequence as a prerequisite, as the transcript sequences can be compared to the closest anno- tated reference sequence in the public database using stan- dard computational tools). 8. The PICR service reconciles protein identifiers across multiple source databases (http:/ /www.ebi.ac.uk/tools/picr).
- 44. 25 Omics Technologies, Dataand Bioinformatics Principles 9. InterPro/InterProScan: InterPro is a database of protein families, domains, regions, repeats and sites in which identifi- able features found in known proteins can be applied to new protein sequences (http:/ /www.ebi.ac.uk/interpro/index. html). InterPro combines a number of databases (referred to as member databases) that use different methodologies and a varying degree of biological information on well-character- ised proteins to derive protein signatures. By uniting the member databases, InterPro capitalises on their individual strengths, producing a powerful integrated database and diagnostic tool: InterProScan. InterProScan is a sequence search package that combines the individual search methods of the member databases and provides the results in a consis- tent format: The user can choose among text, raw, HTML or XML. The results display potential GO terms and the InterPro entry relationships where applicable (http:/ /www.ebi.ac.uk/ Tools/InterProScan). 10. NCBI RefSeq databases: The Reference Sequence (RefSeq) database is a non-redundant collection of richly annotated DNA, RNA, and protein sequences from diverse taxa, see http:/ /www.ncbi.nlm.nih.gov/RefSeq. 11. IntEnz: Integrated relational Enzyme database is a freely available resource focused on enzyme nomenclature (http:/ / www.ebi.ac.uk/intenz). 12. OMIM: the Online Mendelian Inheritance in Man database (http:/ /www.ncbi.nlm.nih.gov/omim). 13. Genomic databases from potential pathogens: EchoBase is a database that curates new experimental and bioinformatic information about the genes and gene products of the model bacterium Escherichia coli K-12 strain MG1655; http:/ /www. york.ac.uk/res/thomas. Ecogene database contains updated information about the E. coli K-12 genome and proteome sequences, including extensive gene bibliographies; http:/ /ecogene.org. LegioList is a database dedicated to the analysis of the genomes of Legionella pneumophila strain Paris (endemic in France), strain Lens (epidemic isolate), strain Philadelphia 1, and strain Corby; http:/ /genolist.pasteur.fr/LegioList. 14. Tranche: Tranche is a free and open source file sharing tool that facilitates the storage of large amounts of data, see https:/ /trancheproject.org. 15. PeptideAtlas: PeptideAtlas (http:/ /www.peptideatlas.org) is a multi-organism, publicly accessible compendium of peptides identified in a large set of tandem mass spectrometry pro- teomics experiments.
- 45. 26 Schneider andOrchard 16. The Global Proteome Machine: Open-source, freely available informatics system for the identification of proteins using tan- dem mass spectra of peptides derived from an enzymatic digest of a mixture of mature proteins, for more see http:/ / www.thegpm.org. 17. EMBOSS: EMBOSS is “The European Molecular Biology Open Software Suite”. It is a free, Open Source software analysis package especially designed for the needs of the molecular biology user community. EMBOSS automatically copes with data in a variety of formats and allows transparent retrieval of sequence data from the web, see http:/ /emboss. sourceforge.net/what. 18. Selected projects, organisations and institutes relevant in Omics http:/ /www.ebi.ac.uk http:/ /www.ncbi.nlm.nih.gov http:/ /www.bii.a-star.edu.sg http:/ /www.ibioinformatics.org http:/ /www.bioinformatics.org.nz http:/ /www.isb-sib.ch http:/ /www.igb.uci.edu http:/ /www.uhnres.utoronto.ca/centres/proteomics http:/ /www.humanvariomeproject.org http:/ /www.expasy.org/links.html http:/ /bioinfo.cipf.es http:/ /www.bcgsc.ca http:/ /www.blueprint.org http:/ /www.cmbi.kun.nl/edu/webtutorials http:/ /newscenter.cancer.gov/sciencebehind http:/ /www.genome.gov/Research http:/ /cmgm.stanford.edu 19. Genomics related resources Genomes Pages at the EBI: http:/ /www.ebi.ac.uk/genomes http:/ /www.ensembl.org/index.html, http:/ /www.ensemblgenomes.org Caenorhabditis elegans (and some other nematodes): http:/ / www.wormbase.org Database for Drosophila melanogaster: http:/ /flybase.org Mouse Genome Informatics: http:/ /www.informatics.jax.org Rat Genome Database: http:/ /rgd.mcw.edu
- 46. 27 Omics Technologies, Dataand Bioinformatics Principles Saccharomyces Genome Database: http:/ /www.yeastgenome.org Pombe genome Project: http:/ /www.sanger.ac.uk/Projects/ S_pombe AceDB genome database: http:/ /www.acedb.org/introduc- tion.shtml HIV Sequence Database: http:/ /www.hiv.lanl.gov/content/ sequence/HIV/mainpage.html 3-D structural information about nucleic acids: http:/ / ndbserver.rutgers.edu Gene Ontology: http:/ /www.geneontology.org Human mitochondrial genome database: http:/ /www. mitomap.org 20. Transcriptomics related resources ArrayExpress: http:/ /www.ebi.ac.uk/microarray-as/ae Gene Expression Omnibus: http:/ /www.ncbi.nlm.nih.gov/geo MGED Society: http:/ /www.mged.org miRBASE: http:/ /www.mirbase.org Comparative RNA: http:/ /www.rna.ccbb.utexas.edu Arabidopsis gene expression database: http:/ /www.arexdb.org Noncoding RNA database: http:/ /www.ncrna.org/frnadb Mammalian noncoding RNA database: http:/ /jsm-research. imb.uq.edu.au/rnadb Noncoding RNA databases: http:/ /biobases.ibch.poznan.pl/ ncRNA Comprehensive Ribosomal RNA database: http:/ /www. arb-silva.de RNA modification pathways: http:/ /modomics.genesilico.pl RNA modification database: http:/ /library.med.utah.edu/ RNAmods RNAi database: http:/ /nematoda.bio.nyu.edu:8001/cgi-bin/ index.cgi Genomic tRNA database: http:/ /gtrnadb.ucsc.edu MicroCosm Targets: http:/ /www.ebi.ac.uk/enright-srv/ microcosm/htdocs/targets/v5 miRNA sequences: http:/ /www.ebi.ac.uk/enright-srv/MapMi microRNA binding and siRNA off-target effects: http:/ / www.ebi.ac.uk/enright/sylamer 21. Proteomics related resources Protein sequences: http:/ /www.uniprot.org ExPaSy Proteomics Service: http:/ /www.expasy.org
- 47. 28 Schneider andOrchard Protein information Resources: http:/ /pir.georgetown.edu Gene Ontology (GO) annotations to proteins: http:/ /www. ebi.ac.uk/GOA/index.html The Peptidase database: http:/ /merops.sanger.ac.uk Molecular Class-Specific Information System (MCSIS) proj- ect: http:/ /www.gpcr.org PROWL (Mass spectrometry and Gaseous Ion Chemistry): http:/ /prowl.rockefeller.edu Protein fingerprinting: http:/ /www.bioinf.manchester.ac.uk/ dbbrowser/PRINTS/index.php Protein families: http:/ /pfam.sanger.ac.uk Domain Prediction: http:/ /hydra.icgeb.trieste.it/~kristian/ SBASE Protein domain families: http:/ /prodom.prabi.fr/prodom/ current/html/home.php Protein families, domains and regions: http:/ /www.ebi.ac. uk/interpro/index.html Simple Modular Architecture Research Tool: http:/ /smart. embl-heidelberg.de Integrated Protein Knowledgebase: http:/ /pir.georgetown. edu/iproclass TIGRFAMS: http:/ /www.jcvi.org/cms/research/projects/ tigrfams/overview Protein databank: http:/ /www.rcsb.org/pdb/home/home.do PRIDE: http:/ /www.ebi.ac.uk/pride Protein Data Bank in Europe: http:/ /www.ebi.ac.uk/pdbe Peptidome: http:/ /www.ncbi.nlm.nih.gov/peptidome PeptideAtlas: http:/ /www.peptideatlas.org Global Proteomics Machine: http:/ /www.thegpm.org/ GPMDB/index.html 22. Protein–protein interaction databases IntAct: http:/ /www.ebi.ac.uk/intact/main.xhtml IMEx: http:/ /imex.sourceforge.net DIP: http:/ /dip.doe-mbi.ucla.edu MINT: http:/ /mint.bio.uniroma2.it/mint MPact: http:/ /mips.gsf.de/genre/proj/mpact MatrixDB: http:/ /matrixdb.ibcp.fr MPIDB: http:/ /www.jcvi.org/mpidb BioGRID: http:/ /www.thebiogrid.org
- 48. 29 Omics Technologies, Dataand Bioinformatics Principles Acknowledgements The authors would like to thank Dr. Gabriella Rustici and Dr. Daniel Zerbino for useful insights and information on tran- scriptomics and genome assembly respectively. The authors would also like to thank Dr. James Watson for useful comments to the manuscript. References 1. Knasmüller, S. et al. (2008) Use of conven- tional and -omics based methods for health claims of dietary antioxidants: A critical over- view. Br J Nutr 99, ES3–52. 2. Hillieret, L.W. et al. (2008) Whole-genome sequencing and variant discovery in C. elegans. Nat Methods 5, 183–88. 3. Johnson, D.S. et al. (2007) Genome-wide mapping of in vivo protein–DNA interactions. Science 316, 1441–42. 4. Mortazavi, A. et al. (2008) Mapping and quantifying mammalian transcriptomes by RNA-Seq. Nat Methods 5, 621–8. 5. Rustici, G. et al. (2008) Data storage and analysis in ArrayExpress and Expression Profiler. Curr Protoc Bioinformatics 7, 7–13. 6. Whetzel, P.L. et al. (2006) The MGED Ontology: A resource for semantics-based description of microarray experiments. Bioinformatics 22, 866–73. 7. Burge, C., Birney, E., and Fickett, J. (2002) Top 10 future challenges for bioinformatics. Genome Technol 17, 1–3. 8. Havlak, P. et al. (2004) The Atlas genome assembly system. Genome Res 14, 721–32. 9. Batzoglou, S. et al. (2002) ARACHNE: A whole genome shotgun assembler. Genome Res 12, 177–89. 10. Myers, E.W. et al. (2000) A whole-genome assembly of Drosophila. Science 287, 2196–204. 11. Huang, X. et al. (2003) PCAP: A whole- genome assembly program. Genome Res 13, 2164–70. 12. Mullikin, J.C., and Ning, Z. (2003) The Phusion assembler. Genome Res 13, 81–90. 13. Pevzner, P.A., Tang, H., and Waterman, M.S. (2001) An Eulerian path approach to DNA fragment assembly. Proc Natl Acad Sci USA 14, 9748–53. 14. Hernandez, D. et al. (2008) De novo bacte- rial genome sequencing: Millions of very short reads assembled on a desktop computer. Genome Res 18, 802–9. 15. Idury, R., and Waterman, M. (1995) A new algorithm for DNA sequence assembly. J Comput Biol 2, 291–306. 16. Pevzner, P., and Tang, H. (2001) Fragment assembly with double-barrelled data. Bioinformatics 17, S225–33. 17. Chaisson, M.J., and Pevzner, P.A. (2008) Short read fragment assembly of bacterial genomes. Genome Res 18, 324–30. 18. Zerbino, D.R., and Birney, E. (2008) Velvet: Algorithms for de novo short read assembly using de Bruijn graphs. Genome Res 18, 821–9. 19. Ossowski, S. et al. (2008) Sequencing of nat- ural strains of Arabidopsis thaliana with short reads. Genome Res 18, 2024–33. 20. Farrer, R.A. et al. (2009) De novo assembly of the Pseudomonas syringae pv. syringae B728a genome using Illumina/Solexa short sequence reads. FEMS Microbiol Lett 1, 103–11. 21. Wakaguri, H. et al. (2008) DBTSS: Database of transcription start sites, progress report. Nucleic Acids Res 36, D97–101. 22. Chen, X. et al. (2009) High throughput genome-wide survey of small RNAs from the parasitic protists Giardia intestinalis and Trichomonas vaginalis. Genome Biol Evol 1, 165–75. 23. Butler, J. et al. (2008) ALLPATHS: De novo assembly of whole-genome shotgun micro- reads. Genome Res 18, 810–20. 24. Chen, J., and Skiena, S. (2007) Assembly for double-ended short-read sequencing technol- ogies. In ‘Advances in Genome Sequencing Technology and Algorithms’, edited by E. Mardis, S. Kim, and H. Tang. Artech House Publishers, Boston. 25. Simpson, J.T. et al. (2009) ABySS: A parallel assembler for short read sequence data. Genome Res 9, 1117–23.
- 49. 30 Schneider andOrchard 26. Jackson, B.G., Schnable, P.S., and Aluru, S. (2009) Parallel short sequence assembly of tran- scriptomes. BMC Bioinformatics 10, S1–14. 27. Spudich, G., Fernandez-Suarez, X.M., and Birney, E. (2007) Genome browsing with Ensembl: A practical overview. Brief Funct Genomic Proteomic 6, 202–19. 28. Vizcaíno, J.A. et al. (2009) A guide to the Proteomics Identifications Database proteom- ics data repository. Proteomics 9, 4276–83. 29. Hunter, S. et al. (2009) InterPro: The inte- grative protein signature database. Nucleic Acids Res 37, 211–15. 30. Cesareni, G. et al. (2005) Comparative inter- atcomics. FEBS Lett 579, 1828–33. 31. Brazma, A. et al. (2001) Minimum informa- tionaboutamicroarrayexperiment(MIAME)- toward standards for microarray data. Nat Genet 29, 365–71. 32. Levy, S. et al. (2007) The diploid genome sequence of an individual human. PLoS Biol 5, 2113–44. 33. Wheeler, D.A. et al. (2008) The complete genome of an individual by massively parallel DNA sequencing. Nature 452, 872–76. 34. Venter, J.C. et al. (2001) The sequence of the human genome. Science 291, 1304–51. 35. Spencer, C.C. et al. (2006) The influence of recombination on human genetic diversity. PLoS Genet 2, e148. 36. Brazma, A. et al. (2003) ArrayExpress – A pub- lic repository for microarray gene expression data at the EBI. Nucleic Acids Res 31, 68–71. 37. Edgar, R., Domrachev, M., and Lash, A.E. (2002) Gene Expression Omnibus: NCBI gene expression and hybridization array data repository. Nucleic Acids Res 30, 207–10. 38. Ikeo, K. et al. (2003) CIBEX: Center for information biology gene expression database. C R Biol 326, 1079–82. 39. Parkinson, H. et al. (2009) ArrayExpress update – From an archive of functional genomics experiments to the atlas of gene expression. Nucleic Acids Res 37, 868–72. 40. Aranda, B. et al. (2009) The IntAct molecular interaction database. Nucleic Acid Res. 1–7 doi:10.1093/nar/gkp878. 41. Orchard, O. et al. (2007) The minimum information required for reporting a molecu- lar interaction experiment (MIMIx). Nat Biotechnol 25, 894–8. 42. Kiemer, L., and Cesareni, G. (2007) Comparative interactomics: Comparing apples and pears? Trends Biotechnol 25, 448–54. 43. Kiemer, L. et al. (2007) WI-PHI: A weighted yeast interactome enriched for direct physical interactions. Proteomics 7, 932–43. 44. Joyce, A.R., and Palsson, B.Ø. (2006) The model organism as a system: Integrating ‘omics’ data sets. Nat Rev Mol Cell Biol 7, 198–210. 45. Akula, S.P. et al. (2009) Techniques for inte- grating -omics Data. Bioinformation 3, 284–6. 46. Haider, S. et al. (2009) BioMart Central Portal – Unified access to biological data. Nucleic Acids Res 1, W23–27. 47. Li, P. et al. (2008) Performing statistical analyses on quantitative data in Taverna work- flows: An example using R and maxdBrowse to identify differentially-expressed genes from microarray data. BMC Bioinformatics 9, 334.
- 50. 31 Bernd Mayer (ed.),Bioinformatics for Omics Data: Methods and Protocols, Methods in Molecular Biology, vol. 719, DOI 10.1007/978-1-61779-027-0_2, © Springer Science+Business Media, LLC 2011 Chapter 2 Data Standards for Omics Data: The Basis of Data Sharing and Reuse Stephen A. Chervitz, Eric W. Deutsch, Dawn Field, Helen Parkinson, John Quackenbush, Phillipe Rocca-Serra, Susanna-Assunta Sansone, Christian J. Stoeckert Jr., Chris F. Taylor, Ronald Taylor, and Catherine A. Ball Abstract To facilitate sharing of Omics data, many groups of scientists have been working to establish the relevant data standards. The main components of data sharing standards are experiment description standards, data exchange standards, terminology standards, and experiment execution standards. Here we provide a survey of existing and emerging standards that are intended to assist the free and open exchange of large-format data. Key words: Data sharing, Data exchange, Data standards, MGED, MIAME, Ontology, Data format, Microarray, Proteomics, Metabolomics The advent of genome sequencing efforts in the 1990s led to a dramatic change in the scale of biomedical experiments. With the comprehensive lists of genes and predicted gene products that resulted from genome sequences, researchers could design exper- iments that assayed every gene, every protein, or every predicted metabolite. When exploiting transformative Omics technologies such as microarrays, proteomics or high-throughput cell assays, a single experiment can generate very large amounts of raw data as well as summaries in the form of lists of sequences, genes, pro- teins, metabolites, or SNPs. Managing, analyzing, and sharing the large data set from Omics experiments present challenges 1. Introduction
- 51. Another Random ScribdDocument with Unrelated Content
- 52. —Ah! mon petit,dit Blondet, je te croyais plus fort! Non, ma parole d'honneur, en regardant ton front, je te douais d'une omnipotence semblable à celle des grands esprits, tous assez puissamment constitués pour pouvoir considérer toute chose dans sa double forme. Mon petit, en littérature, chaque idée a son envers et son endroit; et personne ne peut prendre sur lui d'affirmer quel est l'envers. Tout est bilatéral dans le domaine de la pensée. Les idées son binaires. Janus est le mythe de la critique et le symbole du génie. Il n'y a que Dieu de triangulaire! Ce qui met Molière et Corneille hors ligne, n'est-ce pas la faculté de faire dire oui à Alceste et non à Philinte, à Octave et à Cinna. Rousseau, dans la Nouvelle- Héloïse, a écrit une lettre pour et une lettre contre le duel, oserais-tu prendre sur toi de déterminer sa véritable opinion? Qui de nous pourrait prononcer entre Clarisse et Lovelace, entre Hector et Achille? Quel est le héros d'Homère? quelle fut l'intention de Richardson? La critique doit contempler les œuvres sous tous leurs aspects. Enfin nous sommes de grands rapporteurs. —Vous tenez donc à ce que vous écrivez? lui dit Vernou d'un air railleur. Mais nous sommes des marchands de phrases, et nous vivons de notre commerce. Quand vous voudrez faire une grande et belle œuvre, un livre enfin, vous pourrez y jeter vos pensées, votre âme, vous y attacher, le défendre; mais des articles lus aujourd'hui, oubliés demain, ça ne vaut à mes yeux que ce qu'on les paye. Si vous mettez de l'importance à de pareilles stupidités, vous ferez donc le signe de la croix et vous invoquerez l'Esprit saint pour écrire un prospectus! Tous parurent étonnés de trouver à Lucien des scrupules et achevèrent de mettre en lambeaux sa robe prétexte pour lui passer la robe virile des journalistes. —Sais-tu par quel mot s'est consolé Nathan après avoir lu ton article? dit Lousteau. —Comment le saurais-je?
- 53. —Nathan s'est écrié:—Lespetits articles passent, les grands ouvrages restent! Cet homme viendra souper ici dans deux jours, il doit se prosterner à tes pieds, baiser ton ergot, et te dire que tu es un grand homme. —Ce serait drôle, dit Lucien. —Drôle! reprit Blondet, c'est nécessaire. —Mes amis, je veux bien, dit Lucien un peu gris; mais comment faire? —Eh! bien, dit Lousteau, écris pour le journal de Merlin trois belles colonnes où tu te réfuteras toi-même. Après avoir joui de la fureur de Nathan, nous venons de lui dire qu'il nous devrait bientôt des remercîments pour la polémique serrée à l'aide de laquelle nous allions faire enlever son livre en huit jours. Dans ce moment-ci, tu es, à ses yeux, un espion, une canaille, un drôle; après-demain tu seras un grand homme, une tête forte, un homme de Plutarque! Nathan t'embrassera comme son meilleur ami. Dauriat est venu, tu as trois billets de mille francs: le tour est fait. Maintenant il te faut l'estime et l'amitié de Nathan. Il ne doit y avoir d'attrapé que le libraire. Nous ne devons immoler et poursuivre que nos ennemis. S'il s'agissait d'un homme qui eût conquis un nom sans nous, d'un talent incommode et qu'il fallût annuler, nous ne ferions pas de réplique semblable; mais Nathan est un de nos amis, Blondet l'avait fait attaquer dans le Mercure pour se donner le plaisir de répondre dans les Débats. Aussi la première édition du livre s'est-elle enlevée! —Mes amis, foi d'honnête homme, je suis incapable d'écrire deux mots d'éloge sur ce livre... —Tu auras encore cent francs, dit Merlin, Nathan t'aura déjà rapporté dix louis, sans compter un article que tu peux faire dans la Revue de Finot, et qui te sera payé cent francs par Dauriat et cent francs par la Revue: total, vingt louis! —Mais que dire? demanda Lucien.
- 54. —Voici comment tupeux t'en tirer, mon enfant, répondit Blondet en se recueillant. L'envie, qui s'attache à toutes les belles œuvres, comme le ver aux beaux et bons fruits, a essayé de mordre sur ce livre, diras-tu. Pour y trouver des défauts, la critique a été forcée d'inventer des théories à propos de ce livre, de distinguer deux littératures: celle qui se livre aux idées et celle qui s'adonne aux images. Là, mon petit, tu diras que le dernier degré de l'art littéraire est d'empreindre l'idée dans l'image. En essayant de prouver que l'image est toute la poésie, tu te plaindras du peu de poésie que comporte notre langue, tu parleras des reproches que nous font les étrangers sur le positivisme de notre style, et tu loueras monsieur de Canalis et Nathan des services qu'ils rendent à la France en déprosaïsant son langage. Accable ta précédente argumentation en faisant voir que nous sommes en progrès sur le dix-huitième siècle. Invente le Progrès (une adorable mystification à faire aux bourgeois)! Notre jeune littérature procède par tableaux où se concentrent tous les genres, la comédie et le drame, les descriptions, les caractères, le dialogue, sortis par les nœuds brillants d'une intrigue intéressante. Le roman, qui veut le sentiment, le style et l'image, est la création moderne la plus immense. Il succède à la comédie qui, dans les mœurs modernes, n'est plus possible avec ses vieilles lois; il embrasse le fait et l'idée dans ses inventions qui exigent et l'esprit de La Bruyère et sa morale incisive, les caractères traités comme l'entendait Molière, les grandes machines de Shakspeare et la peinture des nuances les plus délicates de la passion, unique trésor que nous aient laissé nos devanciers. Aussi le roman est-il bien supérieur à la discussion froide et mathématique, à la sèche analyse du dix-huitième siècle. Le roman, diras-tu sentencieusement, est une épopée amusante. Cite Corinne, appuie-toi sur madame de Staël. Le dix-huitième siècle a tout mis en question, le dix-neuvième est chargé de conclure: aussi conclut-il par des réalités; mais par des réalités qui vivent et qui marchent; enfin il met en jeu la passion, élément inconnu à Voltaire. Tirade contre Voltaire. Quant à Rousseau, il n'a fait qu'habiller des raisonnements et des systèmes. Julie et Claire sont des entéléchies, elles n'ont ni chair ni os. Tu peux démancher sur ce thème et dire
- 55. que nous devonsà la paix, aux Bourbons, une littérature jeune et originale, car tu écris dans un journal Centre droit. Moque-toi des faiseurs de systèmes. Enfin tu peux t'écrier par un beau mouvement: Voilà bien des erreurs, bien des mensonges chez notre confrère! et pourquoi? pour déprécier une belle œuvre, tromper le public et arriver à cette conclusion: Un livre qui se vend ne se vend pas. Proh pudor! lâche Proh pudor! ce juron honnête anime le lecteur. Enfin annonce la décadence de la critique! Conclusion: Il n'y a qu'une seule littérature, celle des livres amusants. Nathan est entré dans une voie nouvelle, il a compris son époque et répond à ses besoins. Le besoin de l'époque est le drame. Le drame est le vœu du siècle où la politique est un mimodrame perpétuel. N'avons-nous pas vu en vingt ans, diras-tu, les quatre drames de la Révolution, du Directoire, de l'Empire et de la Restauration? De là, tu roules dans le dithyrambe de l'éloge, et la seconde édition s'enlève; car, samedi prochain, tu feras une feuille dans notre Revue, et tu la signeras de Rubempré en toutes lettres. Dans ce dernier article, tu diras: Le propre des belles œuvres est de soulever d'amples discussions. Cette semaine tel journal a dit telle chose du livre de Nathan, tel autre lui a vigoureusement répondu. Tu critiques les deux critiques C. et L., tu me dis en passant une politesse à propos de mon article des Débats, et tu finis en affirmant que l'œuvre de Nathan est le plus beau livre de l'époque. C'est comme si tu ne disais rien, on dit cela de tous les livres. Tu auras gagné quatre cents francs dans ta semaine, outre le plaisir d'écrire la vérité quelque part. Les gens sensés donneront raison ou à C. ou à L. ou à Rubempré, peut-être à tous trois! La mythologie, qui certes est une des plus grandes inventions humaines, a mis la Vérité dans le fond d'un puits, ne faut-il pas des seaux pour l'en tirer? tu en auras donné trois pour un au public? Voilà, mon enfant. Marche! Lucien fut étourdi, Blondet l'embrassa sur les deux joues en lui disant:—Je vais à ma boutique. Chacun s'en alla à sa boutique; car, pour ces hommes forts, le journal était une boutique. Tous devaient se revoir le soir aux Galeries-de-Bois, où Lucien irait signer son traité chez Dauriat. Florine et Lousteau, Lucien et Coralie, Blondet et Finot dînaient au
- 56. Palais-Royal, où DuBruel traitait le directeur du Panorama- Dramatique. —Ils ont raison! s'écria Lucien quand il fut seul avec Coralie, les hommes doivent être des moyens entre les mains des gens forts. Quatre cents francs pour trois articles! Doguereau me les donnait à peine pour un livre qui m'a coûté deux ans de travail. —Fais de la critique, dit Coralie, amuse-toi! Est-ce que je ne suis pas ce soir en Andalouse, demain ne me mettrai-je pas en bohémienne, un autre jour en homme? Fais comme moi, donne-leur des grimaces pour leur argent, et vivons heureux. Lucien, épris du paradoxe, fit monter son esprit sur ce mulet capricieux, fils de Pégase et de l'ânesse de Balaam. Il se mit à galoper dans les champs de la pensée pendant sa promenade au Bois, et découvrit des beautés originales dans la thèse de Blondet. Il dîna comme dînent les gens heureux, il signa chez Dauriat un traité par lequel il lui cédait en toute propriété le manuscrit des Marguerites sans y apercevoir aucun inconvénient; puis il alla faire un tour au journal, où il brocha deux colonnes, et revint rue de Vendôme. Le lendemain matin, il se trouva que les idées de la veille avaient germé dans sa tête, comme il arrive chez tous les esprits pleins de séve dont les facultés ont encore peu servi. Lucien éprouva du plaisir à méditer ce nouvel article, il s'y mit avec ardeur. Sous sa plume se rencontrèrent les beautés que fait naître la contradiction. Il fut spirituel et moqueur, il s'éleva même à des considérations neuves sur le sentiment et l'image en littérature. Ingénieux et fin, il retrouva, pour louer Nathan, ses premières impressions à la lecture du livre au cabinet littéraire de la cour du Commerce. De sanglant et âpre critique, de moqueur comique, il devint poète en quelques phrases finales qui se balancèrent majestueusement comme un encensoir chargé de parfums vers l'autel. —Cent francs, Coralie! dit-il en montrant les huit feuillets de papier écrits pendant qu'elle s'habillait.
- 57. Dans la verveoù il était, il fit à petites plumées l'article terrible promis à Blondet contre Châtelet et madame de Bargeton. Il goûta pendant cette matinée l'un des plaisirs secrets les plus vifs des journalistes, celui d'aiguiser l'épigramme, d'en polir la lame froide qui trouve sa gaîne dans le cœur de la victime, et de sculpter le manche pour les lecteurs. Le public admire le travail spirituel de cette poignée, il n'y entend pas malice, il ignore que l'acier du bon mot altéré de vengeance barbote dans un amour-propre fouillé savamment, blessé de mille coups. Cet horrible plaisir, sombre et solitaire, dégusté sans témoins, est comme un duel avec un absent, tué à distance avec le tuyau d'une plume, comme si le journaliste avait la puissance fantastique accordée aux désirs de ceux qui possèdent des talismans dans les contes arabes. L'épigramme est l'esprit de la haine, de la haine qui hérite de toutes les mauvaises passions de l'homme, de même que l'amour concentre toutes ses bonnes qualités. Aussi n'est-il pas d'homme qui ne soit spirituel en se vengeant, par la raison qu'il n'en est pas un à qui l'amour ne donne des jouissances. Malgré la facilité, la vulgarité de cet esprit en France, il est toujours bien accueilli. L'article de Lucien devait mettre et mit le comble à la réputation de malice et de méchanceté du journal; il entra jusqu'au fond de deux cœurs, il blessa grièvement madame de Bargeton, son ex-Laure, et le baron Châtelet, son rival. —Eh! bien, allons faire une promenade au Bois, les chevaux sont mis, et ils piaffent, lui dit Coralie; il ne faut pas se tuer. —Portons l'article sur Nathan chez Hector. Décidément le journal est comme la lance d'Achille qui guérissait les blessures qu'elle avait faites, dit Lucien en corrigeant quelques expressions. Les deux amants partirent et se montrèrent dans leur splendeur à ce Paris qui, naguère, avait renié Lucien, et qui maintenant commençait à s'en occuper. Occuper Paris de soi quand on a compris l'immensité de cette ville et la difficulté d'y être quelque chose, causa d'enivrantes jouissances qui grisèrent Lucien. —Mon petit, dit l'actrice, passons chez ton tailleur presser tes habits ou les essayer s'ils sont prêts. Si tu vas chez tes belles
- 58. madames, je veuxque tu effaces ce monstre de De Marsay, le petit Rastignac, les Ajuda-Pinto, les Maxime de Trailles, les Vandenesse, enfin tous les élégants. Songe que ta maîtresse est Coralie! Mais ne me fais pas de traits, hein? Deux jours après, la veille du souper offert par Lucien et Coralie à leurs amis, l'Ambigu donnait une pièce nouvelle dont le compte devait être rendu par Lucien. Après leur dîner, Lucien et Coralie allèrent à pied de la rue de Vendôme au Panorama-Dramatique, par le boulevard du Temple du côté du café Turc, qui, dans ce temps-là, était un lieu de promenade en faveur. Lucien entendit vanter son bonheur et la beauté de sa maîtresse. Les uns disaient que Coralie était la plus belle femme de Paris, les autres trouvaient Lucien digne d'elle. Le poète se sentit dans son milieu. Cette vie était sa vie. Le Cénacle, à peine l'apercevait-il. Ces grands esprits qu'il admirait tant deux mois auparavant, il se demandait s'ils n'étaient pas un peu niais avec leurs idées et leur puritanisme. Le mot de jobards, dit insouciamment par Coralie, avait germé dans l'esprit de Lucien, et portait déjà ses fruits. Il mit Coralie dans sa loge, flâna dans les coulisses du théâtre où il se promenait en sultan, où toutes les actrices le caressaient par des regards brûlants et par des mots flatteurs. —Il faut que j'aille à l'Ambigu faire mon métier, dit-il. A l'Ambigu, la salle était pleine. Il ne s'y trouva pas de place pour Lucien. Lucien alla dans les coulisses et se plaignit amèrement de ne pas être placé. Le régisseur, qui ne le connaissait pas encore, lui dit qu'on avait envoyé deux loges à son journal, et l'envoya promener. —Je parlerai de la pièce selon ce que j'en aurai entendu, dit Lucien d'un air piqué. —Êtes-vous bête? dit la jeune première au régisseur, c'est l'amant de Coralie! Aussitôt le régisseur se retourna vers Lucien et lui dit:—Monsieur, je vais aller parler au directeur.
- 59. Ainsi les moindresdétails prouvaient à Lucien l'immensité du pouvoir du journal et caressaient sa vanité. Le directeur vint et obtint du duc de Rhétoré et de Tullia, le premier sujet, qui se trouvaient dans une loge d'avant-scène, de prendre Lucien avec eux. Le duc y consentit en reconnaissant Lucien. —Vous avez réduit deux personnes au désespoir, lui dit le jeune homme en lui parlant du baron Châtelet et de madame de Bargeton. —Que sera-ce donc demain? dit Lucien. Jusqu'à présent mes amis se sont portés contre eux en voltigeurs, mais je tire à boulet rouge cette nuit. Demain, vous verrez pourquoi nous nous moquons de Potelet. L'article est intitulé: Potelet de 1811 à Potelet de 1821. Châtelet sera le type des gens qui ont renié leur bienfaiteur en se ralliant aux Bourbons. Après avoir fait sentir tout ce que je puis, j'irai chez madame de Montcornet. Lucien eut avec le jeune duc une conversation étincelante d'esprit; il était jaloux de prouver à ce grand seigneur combien mesdames d'Espard et de Bargeton s'étaient grossièrement trompées en le méprisant; mais il montra le bout de l'oreille en essayant d'établir ses droits à porter le nom de Rubempré, quand, par malice, le duc de Rhétoré l'appela Chardon. —Vous devriez, lui dit le duc, vous faire royaliste. Vous vous êtes montré homme d'esprit, soyez maintenant homme de bon sens. La seule manière d'obtenir une ordonnance du roi qui vous rende le titre et le nom de vos ancêtres maternels, est de la demander en récompense des services que vous rendrez au Château. Les Libéraux ne vous feront jamais comte! Voyez-vous, la Restauration finira par avoir raison de la Presse, la seule puissance à craindre. On a déjà trop attendu, elle devrait être muselée. Profitez de ses derniers moments de liberté pour vous rendre redoutable. Dans quelques années, un nom et un titre seront en France des richesses plus sûres que le talent. Vous pouvez ainsi tout avoir: esprit, noblesse et beauté, vous arriverez à tout. Ne soyez donc en ce moment libéral que pour vendre avec avantage votre royalisme.
- 60. Le duc priaLucien d'accepter l'invitation à dîner que devait lui envoyer le ministre avec lequel il avait soupé chez Florine. Lucien fut en un moment séduit par les réflexions du gentilhomme, et charmé de voir s'ouvrir devant lui les portes des salons d'où il se croyait à jamais banni quelques mois auparavant. Il admira le pouvoir de la pensée. La Presse et l'esprit étaient donc le moyen de la société présente. Lucien comprit que peut-être Lousteau se repentait de lui avoir ouvert les portes du temple, il sentait déjà pour son propre compte la nécessité d'opposer des barrières difficiles à franchir aux ambitions de ceux qui s'élançaient de la province vers Paris. Un poète serait venu vers lui comme il s'était jeté dans les bras d'Étienne, il n'osait se demander quel accueil il lui ferait. Le jeune duc aperçut chez Lucien les traces d'une méditation profonde et ne se trompa point en en cherchant la cause: il avait découvert à cet ambitieux, sans volonté fixe, mais non sans désir, tout l'horizon politique comme les journalistes lui avaient montré en haut du Temple, ainsi que le démon à Jésus, le monde littéraire et ses richesses. Lucien ignorait la petite conspiration ourdie contre lui par les gens que blessait en ce moment le journal, et dans laquelle monsieur de Rhétoré trempait. Le jeune duc avait effrayé la société de madame d'Espard en leur parlant de l'esprit de Lucien. Chargé par madame de Bargeton de sonder le journaliste, il avait espéré le rencontrer à l'Ambigu-Comique. Ni le monde, ni les journalistes n'étaient profonds, ne croyez pas à des trahisons ourdies. Ni l'un ni les autres ils n'arrêtent de plan; leur machiavélisme va pour ainsi dire au jour le jour, et consiste à toujours être là, prêts à tout, prêts à profiter du mal comme du bien, à épier les moments où la passion leur livre un homme. Pendant le souper de Florine, le jeune duc avait reconnu le caractère de Lucien, il venait de le prendre par ses vanités, et s'essayait sur lui à devenir diplomate. Lucien, la pièce jouée, courut à la rue Saint-Fiacre y faire son article sur la pièce. Sa critique fut, par calcul, âpre et mordante; il se plut à essayer son pouvoir. Le mélodrame valait mieux que celui du Panorama-Dramatique; mais il voulait savoir s'il pouvait, comme on le lui avait dit, tuer une bonne et faire réussir une mauvaise pièce.
- 61. Le lendemain, endéjeunant avec Coralie, il déplia le journal, après lui avoir dit qu'il y éreintait l'Ambigu-Comique. Lucien ne fut pas médiocrement étonné de lire, après son article sur madame de Bargeton et sur Châtelet, un compte-rendu de l'Ambigu si bien édulcoré durant la nuit, que, tout en conservant sa spirituelle analyse, il en sortait une conclusion favorable. La pièce devait remplir la caisse du théâtre. Sa fureur ne saurait se décrire; il se proposa de dire deux mots à Lousteau. Il se croyait déjà nécessaire, et se promettait de ne pas se laisser dominer, exploiter comme un niais. Pour établir définitivement sa puissance, il écrivit l'article où il résumait et balançait toutes les opinions émises à propos du livre de Nathan pour la Revue de Dauriat et de Finot. Puis, une fois monté, il brocha l'un de ses articles Variétés dus au petit journal. Dans leur première effervescence, les jeunes journalistes pondent des articles avec amour et livrent ainsi très-imprudemment toutes leurs fleurs. Le directeur du Panorama-Dramatique donnait la première représentation d'un vaudeville, afin de laisser à Florine et à Coralie leur soirée. On devait jouer avant le souper. Lousteau vint chercher l'article de Lucien, fait d'avance sur cette petite pièce, dont il avait vu la répétition générale, afin de n'avoir aucune inquiétude relativement à la composition du numéro. Quand Lucien lui eut lu l'un de ces petits charmants articles sur les particularités parisiennes, qui firent la fortune du journal, Étienne l'embrassa sur les deux yeux et le nomma la providence des journaux. —Pourquoi donc t'amuses-tu à changer l'esprit de mes articles? dit Lucien, qui n'avait fait ce brillant article que pour donner plus de force à ses griefs. —Moi! s'écria Lousteau. —Eh! bien, qui donc a changé mon article? —Mon cher, répondit Étienne en riant, tu n'es pas encore au courant des affaires. L'Ambigu nous prend vingt abonnements, dont neuf seulement sont servis au directeur, au chef d'orchestre, au régisseur, à leurs maîtresses et à trois copropriétaires du théâtre. Chacun des théâtres du boulevard paye ainsi huit cents francs au
- 62. journal. Il ya pour tout autant d'argent en loges données à Finot, sans compter les abonnements des acteurs et des auteurs. Le drôle se fait donc huit mille francs aux boulevards. Par les petits théâtres, juge des grands! Comprends-tu? Nous sommes tenus à beaucoup d'indulgence. —Je comprends que je ne suis pas libre d'écrire ce que je pense.... —Eh! que t'importe, si tu y fais tes orges, s'écria Lousteau. D'ailleurs, mon cher, quel grief as-tu contre le théâtre? il te faut une raison pour échiner la pièce d'hier. Échiner pour échiner, nous compromettrions le journal. Quand le journal frapperait avec justice, il ne produirait plus aucun effet. Le directeur t'a-t-il manqué? —Il ne m'avait pas réservé de place. —Bon, fit Lousteau. Je montrerai ton article au directeur, je lui dirai que je t'ai adouci, tu t'en trouveras mieux que de l'avoir fait paraître. Demande-lui demain des billets, il t'en signera quarante en blanc tous les mois, et je te mènerai chez un homme avec qui tu t'entendras pour les placer; il te les achètera tous à cinquante pour cent de remise sur le prix des places. On fait sur les billets de spectacle le même trafic que sur les livres. Tu verras un autre Barbet, un chef de claque, il ne demeure pas loin d'ici, nous avons le temps, viens? —Mais, mon cher, Finot fait un infâme métier à lever ainsi sur les champs de la pensée des contributions indirectes. Tôt ou tard... —Ah! çà, d'où viens-tu? s'écria Lousteau. Pour qui prends-tu Finot? Sous sa fausse bonhomie, sous cet air Turcaret, sous son ignorance et sa bêtise, il y a toute la finesse du marchand de chapeaux dont il est issu. N'as-tu pas vu dans sa cage, au Bureau du journal, un vieux soldat de l'Empire, l'oncle de Finot? Cet oncle est non-seulement un honnête homme, mais il a le bonheur de passer pour un niais. Il est l'homme compromis dans toutes les transactions pécuniaires. A Paris, un ambitieux est bien riche quand il a près de lui une créature qui consent à être compromise. Il est en politique
- 63. comme en journalismeune foule de cas où les chefs ne doivent jamais être mis en cause. Si Finot devenait un personnage politique, son oncle deviendrait son secrétaire et recevrait pour son compte les contributions qui se lèvent dans les bureaux sur les grandes affaires. Giroudeau, qu'au premier abord on prendrait pour un niais, a précisément assez de finesse pour être un compère indéchiffrable. Il est en vedette pour empêcher que nous ne soyons assommés par les criailleries, par les débutants, par les réclamations, et je ne crois pas qu'il y ait son pareil dans un autre journal. —Il joue bien son rôle, dit Lucien, je l'ai vu à l'œuvre. Étienne et Lucien allèrent dans la rue du Faubourg-du-Temple, où le rédacteur en chef s'arrêta devant une maison de belle apparence. —Monsieur Braulard y est-il? demanda-t-il au portier. —Comment monsieur? dit Lucien. Le chef des claqueurs est donc monsieur? —Mon cher, Braulard a vingt mille livres de rente, il a la griffe des auteurs dramatiques du boulevard qui tous ont un compte courant chez lui, comme chez un banquier. Les billets d'auteur et de faveur se vendent. Cette marchandise, Braulard la place. Fais un peu de statistique, science assez utile quand on n'en abuse pas. A cinquante billets de faveur par soirée à chaque spectacle, tu trouveras deux cent cinquante billets par jour; si, l'un dans l'autre, ils valent quarante sous, Braulard paye cent vingt-cinq francs par jour aux auteurs et court la chance d'en gagner autant. Ainsi, les seuls billets des auteurs lui procurent près de quatre mille francs par mois, au total quarante-huit mille francs par an. Suppose vingt mille francs de perte, car il ne peut pas toujours placer ses billets. —Pourquoi? —Ah! les gens qui viennent payer leurs places au bureau passent concurremment avec les billets de faveur qui n'ont pas de places réservées. Enfin le théâtre garde ses droits de location. Il y a les jours de beau temps, et de mauvais spectacles. Ainsi, Braulard gagne peut-être trente mille francs par an sur cet article. Puis il a ses
- 64. claqueurs, autre industrie.Florine et Coralie sont ses tributaires; si elles ne le subventionnaient pas, elles ne seraient point applaudies à toutes les entrées et leurs sorties. Lousteau donnait cette explication à voix basse en montant l'escalier.
- 65. IMP. E. MARTINET. Braularda vingt mille livres de rentes, puis il a ses claqueurs. (ILLUSIONS PERDUES.)
- 66. —Paris est unsingulier pays, dit Lucien en trouvant l'intérêt accroupi dans tous les coins. Une servante proprette introduisit les deux journalistes chez monsieur Braulard. Le marchand de billets, qui siégeait sur un fauteuil de cabinet, devant un grand secrétaire à cylindre, se leva en voyant Lousteau. Braulard, enveloppé d'une redingote de molleton gris, portait un pantalon à pied et des pantoufles rouges absolument comme un médecin ou comme un avoué. Lucien vit en lui l'homme du peuple enrichi: un visage commun, des yeux gris pleins de finesse, des mains de claqueur, un teint sur lequel les orgies avaient passé comme la pluie sur les toits, des cheveux grisonnants, et une voix assez étouffée. —Vous venez, sans doute, pour mademoiselle Florine, et monsieur pour mademoiselle Coralie, dit-il, je vous connais bien. Soyez tranquille, monsieur, dit-il à Lucien, j'achète la clientèle du Gymnase, je soignerai votre maîtresse et je l'avertirai des farces qu'on voudrait lui faire. —Ce n'est pas de refus, mon cher Braulard, dit Lousteau; mais nous venons pour les billets du journal à tous les théâtres des boulevards: moi comme rédacteur en chef, monsieur comme rédacteur de chaque théâtre. —Ah, oui, Finot a vendu son journal. J'ai su l'affaire. Il va bien, Finot. Je lui donne à dîner à la fin de la semaine. Si vous voulez me faire l'honneur et le plaisir de venir, vous pouvez amener vos épouses, il y aura noces et festins, nous avons Adèle Dupuis, Ducange, Frédéric Du Petit-Méré, mademoiselle Millot ma maîtresse, nous rirons bien! nous boirons mieux! —Il doit être gêné, Ducange, il a perdu son procès. —Je lui ai prêté dix mille francs, le succès de Calas va me les rendre; aussi l'ai-je chauffé! Ducange est un homme d'esprit, il a des moyens... Lucien croyait rêver en entendant cet homme apprécier les talents des auteurs.—Coralie a gagné, lui dit Braulard de l'air d'un juge compétent. Si elle est bonne enfant, je la soutiendrai É
- 67. secrètement contre lacabale à son début au Gymnase. Écoutez? Pour elle, j'aurai des hommes bien mis aux galeries qui souriront et qui feront de petits murmures afin d'entraîner l'applaudissement. Voilà un manége qui pose une femme. Elle me plaît, Coralie, et vous devez être content d'elle, elle a des sentiments. Ah! je puis faire chuter qui je veux... —Mais pour les billets? dit Lousteau. —Hé! bien, j'irai les prendre chez monsieur, vers les premiers jours de chaque mois. Monsieur est votre ami, je le traiterai comme vous. Vous avez cinq théâtres, on vous donnera trente billets; ce sera quelque chose comme soixante-quinze francs par mois. Peut- être désirez-vous une avance? dit le marchand de billets en revenant à son secrétaire et tirant sa caisse pleine d'écus. —Non, non, dit Lousteau, nous garderons cette ressource pour les mauvais jours... —Monsieur, reprit Braulard en s'adressant à Lucien, j'irai travailler avec Coralie ces jours-ci, nous nous entendrons bien. Lucien ne regardait pas sans un étonnement profond le cabinet de Braulard où il voyait une bibliothèque, des gravures, un meuble convenable. En passant par le salon, il en remarqua l'ameublement également éloigné de la mesquinerie et du trop grand luxe. La salle à manger lui parut être la pièce la mieux tenue, il en plaisanta. —Mais Braulard est gastronome, dit Lousteau. Ses dîners, cités dans la littérature dramatique, sont en harmonie avec sa caisse. —J'ai de bons vins, répondit modestement Braulard. Allons, voilà mes allumeurs, s'écria-t-il en entendant des voix enrouées et le bruit de pas singuliers dans l'escalier. En sortant, Lucien vit défiler devant lui la puante escouade des claqueurs et des vendeurs de billets, tous gens à casquettes, à pantalons mûrs, à redingotes râpées, à figures patibulaires, bleuâtres, verdâtres, boueuses, rabougries, à barbes longues, aux yeux féroces et patelins tout à la fois, horrible population qui vit et
- 68. foisonne sur lesboulevards de Paris, qui, le matin, vend des chaînes de sûreté, des bijoux en or pour vingt-cinq sous, et qui claque sous les lustres le soir, qui se plie enfin à toutes les fangeuses nécessités de Paris. —Voilà les Romains! dit Lousteau en riant, voilà la gloire des actrices et des auteurs dramatiques. Vu de près, ça n'est pas plus beau que la nôtre. —Il est difficile, répondit Lucien en revenant chez lui, d'avoir des illusions sur quelque chose à Paris. Il y a des impôts sur tout, on y vend tout, on y fabrique tout, même le succès. Les convives de Lucien étaient Dauriat, le directeur du Panorama, Matifat et Florine, Camusot, Lousteau, Finot, Nathan, Hector Merlin et madame du Val-Noble, Félicien Vernou, Blondet, Vignon, Philippe Bridau, Mariette, Giroudeau, Cardot et Florentine, Bixiou. Il avait invité ses amis du Cénacle. Tullia la danseuse, qui, disait-on, était peu cruelle pour du Bruel, fut aussi de la partie, mais sans son duc, ainsi que les propriétaires des journaux où travaillaient Nathan, Merlin, Vignon et Vernou. Les convives formaient une assemblée de trente personnes, la salle à manger de Coralie ne pouvait en contenir davantage. Vers huit heures, au feu des lustres allumés, les meubles, les tentures, les fleurs de ce logis prirent cet air de fête qui prête au luxe parisien l'apparence d'un rêve. Lucien éprouva le plus indéfinissable mouvement de bonheur, de vanité satisfaite et d'espérance en se voyant le maître de ces lieux, il ne s'expliquait plus ni comment ni par qui ce coup de baguette avait été frappé. Florine et Coralie, mises avec la folle recherche et la magnificence artiste des actrices, souriaient au poète de province comme deux anges chargés de lui ouvrir les portes du palais des Songes. Lucien songeait presque. En quelques mois sa vie avait si brusquement changé d'aspect, il était si promptement passé de l'extrême misère à l'extrême opulence, que par moments il lui prenait des inquiétudes comme aux gens qui, tout en rêvant, se savent endormis. Son œil exprimait néanmoins à la vue de cette belle réalité une confiance à
- 69. laquelle des envieuxeussent donné le nom de fatuité. Lui-même, il avait changé. Heureux tous les jours, ses couleurs avaient pâli, son regard était trempé des moites expressions de la langueur; enfin, selon le mot de madame d'Espard, il avait l'air aimé. Sa beauté y gagnait. La conscience de son pouvoir et de sa force perçait dans sa physionomie éclairée par l'amour et par l'expérience. Il contemplait enfin le monde littéraire et la société face à face, en croyant pouvoir s'y promener en dominateur. A ce poète, qui ne devait réfléchir que sous le poids du malheur, le présent parut être sans soucis. Le succès enflait les voiles de son esquif, il avait à ses ordres les instruments nécessaires à ses projets: une maison montée, une maîtresse que tout Paris lui enviait, un équipage, enfin des sommes incalculables dans son écritoire. Son âme, son cœur et son esprit s'étaient également métamorphosés: il ne songeait plus à discuter les moyens en présence de si beaux résultats. Ce train de maison semblera si justement suspect aux économistes qui ont pratiqué la vie parisienne, qu'il n'est pas inutile de montrer la base, quelque frêle qu'elle fût, sur laquelle reposait le bonheur matériel de l'actrice et de son poète. Sans se compromettre, Camusot avait engagé les fournisseurs de Coralie à lui faire crédit pendant au moins trois mois. Les chevaux, les gens, tout devait donc aller comme par enchantement pour ces deux enfants empressés de jouir, et qui jouissaient de tout avec délices. Coralie vint prendre Lucien par la main et l'initia par avance au coup de théâtre de la salle à manger, parée de son couvert splendide, de ses candélabres chargés de quarante bougies, aux recherches royales du dessert, et au menu, l'œuvre de Chevet. Lucien baisa Coralie au front en la pressant sur son cœur. —J'arriverai, mon enfant, lui dit-il, et je te récompenserai de tant d'amour et de tant de dévouement. —Bah! dit-elle, es-tu content? —Je serais bien difficile. —Eh! bien, ce sourire paye tout, répondit-elle en apportant par un mouvement de serpent ses lèvres aux lèvres de Lucien.
- 70. Ils trouvèrent Florine,Lousteau, Matifat et Camusot en train d'arranger les tables de jeu. Les amis de Lucien arrivaient. Tous ces gens s'intitulaient déjà les amis de Lucien. On joua de neuf heures à minuit. Heureusement pour lui, Lucien ne savait aucun jeu; mais Lousteau perdit mille francs et les emprunta à Lucien qui ne crut pas pouvoir se dispenser de les prêter, car son ami les lui demanda. A dix heures environ, Michel, Fulgence et Joseph se présentèrent. Lucien, qui alla causer avec eux dans un coin, trouva leurs visages assez froids et sérieux, pour ne pas dire contraints. D'Arthez n'avait pu venir, il achevait son livre. Léon Giraud était occupé par la publication du premier numéro de sa Revue. Le Cénacle avait envoyé ses trois artistes qui devaient se trouver moins dépaysés que les autres au milieu d'une orgie. —Eh! bien, mes enfants, dit Lucien en affichant un petit ton de supériorité, vous verrez que le petit farceur peut devenir un grand politique. —Je ne demande pas mieux que de m'être trompé, dit Michel. —Tu vis avec Coralie en attendant mieux? lui demanda Fulgence. —Oui, reprit Lucien d'un air qu'il voulait rendre naïf. Coralie avait un pauvre vieux négociant qui l'adorait, elle l'a mis à la porte. Je suis plus heureux que ton frère Philippe qui ne sait comment gouverner Mariette, ajouta-t-il en regardant Joseph Bridau. —Enfin, dit Fulgence, tu es maintenant un homme comme un autre, tu feras ton chemin. —Un homme qui pour vous restera le même en quelque situation qu'il se trouve, répondit Lucien. Michel et Fulgence se regardèrent en échangeant un sourire moqueur que vit Lucien, et qui lui fit comprendre le ridicule de sa phrase. —Coralie est bien admirablement belle, s'écria Joseph Bridau. Quel magnifique portrait à faire!
- 71. —Et bonne, réponditLucien. Foi d'homme, elle est angélique; mais tu feras son portrait; prends-la, si tu veux, pour modèle de ta Vénitienne amenée au vieillard. —Toutes les femmes qui aiment sont angéliques, dit Michel Chrestien. En ce moment Raoul Nathan se précipita sur Lucien avec une furie d'amitié, lui prit les mains et les lui serra. —Mon bon ami, non-seulement vous êtes un grand homme, mais encore vous avez du cœur, ce qui est aujourd'hui plus rare que le génie, dit-il. Vous êtes dévoué à vos amis. Enfin, je suis à vous à la vie, à la mort, et n'oublierai jamais ce que vous avez fait cette semaine pour moi. Lucien, au comble de la joie en se voyant pateliné par un homme dont s'occupait la Renommée, regarda ses trois amis du Cénacle avec une sorte de supériorité. Cette entrée de Nathan était due à la communication que Merlin lui avait faite de l'épreuve de l'article en faveur de son livre, et qui paraissait dans le journal du lendemain. —Je n'ai consenti à écrire l'attaque, répondit Lucien à l'oreille de Nathan, qu'à la condition d'y répondre moi-même. Je suis des vôtres. Il revint à ses trois amis du Cénacle, enchanté d'une circonstance qui justifiait la phrase de laquelle avait ri Fulgence. —Vienne le livre de d'Arthez, et je suis en position de lui être utile. Cette chance seule m'engagerait à rester dans les journaux. —Y es-tu libre? dit Michel. —Autant qu'on peut l'être quand on est indispensable, répondit Lucien avec une fausse modestie. Vers minuit, les convives furent attablés, et l'orgie commença. Les discours furent plus libres chez Lucien que chez Matifat, car personne ne soupçonna la divergence de sentiments qui existait entre les trois députés du Cénacle et les représentants des journaux.
- 72. Ces jeunes esprits,si dépravés par l'habitude du Pour et du Contre, en vinrent aux prises, et se renvoyèrent les plus terribles axiomes de la jurisprudence qu'enfantait alors le journalisme. Claude Vignon, qui voulait conserver à la critique un caractère auguste, s'éleva contre la tendance des petits journaux vers la personnalité, disant que plus tard les écrivains arriveraient à se déconsidérer eux-mêmes. Lousteau, Merlin et Finot prirent alors ouvertement la défense de ce système, appelé dans l'argot du journalisme la blague, en soutenant que ce serait comme un poinçon à l'aide duquel on marquerait le talent. —Tous ceux qui résisteront à cette épreuve seront des hommes réellement forts, dit Lousteau. —D'ailleurs, s'écria Merlin, pendant les ovations des grands hommes, il faut autour d'eux, comme autour des triomphateurs romains, un concert d'injures. —Eh! dit Lucien, tous ceux de qui l'on se moquera croiront à leur triomphe! —Ne dirait-on pas que cela te regarde? s'écria Finot. —Et nos sonnets! dit Michel Chrestien, ne nous vaudraient-ils pas le triomphe de Pétrarque? —L'or (Laure) y est déjà pour quelque chose, dit Dauriat dont le calembour excita des acclamations générales. —Faciamus experimentum in anima vili, répondit Lucien en souriant. —Eh! malheur à ceux que le Journal ne discutera pas, et auxquels il jettera des couronnes à leur début! Ceux-là seront relégués comme des saints dans leur niche, et personne n'y fera plus la moindre attention, dit Vernou. —On leur dira comme Champcenetz au marquis de Genlis, qui regardait trop amoureusement sa femme:—Passez, bonhomme, on vous a déjà donné, dit Blondet.
- 73. Welcome to ourwebsite – the ideal destination for book lovers and knowledge seekers. With a mission to inspire endlessly, we offer a vast collection of books, ranging from classic literary works to specialized publications, self-development books, and children's literature. Each book is a new journey of discovery, expanding knowledge and enriching the soul of the reade Our website is not just a platform for buying books, but a bridge connecting readers to the timeless values of culture and wisdom. With an elegant, user-friendly interface and an intelligent search system, we are committed to providing a quick and convenient shopping experience. Additionally, our special promotions and home delivery services ensure that you save time and fully enjoy the joy of reading. Let us accompany you on the journey of exploring knowledge and personal growth! ebookultra.com