Building an informatics solution to sustain AI-guided cell profiling with hig...Ola Spjuth
Presentation at SLAS Europe 2019 in Barcelona on 28 june, 2019.
High-content microscopy in automated laboratories present many challenges for storing and processing data, and to build AI models to aid decision making. We have established an informatics system to serve a robotized cell profiling setup with incubators, liquid handling and high-content microscopy for microplates. The informatics system consists of computational infrastructure (CPUs, GPUs, storage), middleware (Kubernetes), imaging database and software (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, and automate the process from acquired images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on data from Cell Painting experiments. The microservice architecture makes the system scalable and expandable, and a key objective is on improving screening and toxicity assessment using AI-aided intelligent experimental design.
Building an informatics solution to sustain AI-guided cell profiling with hig...Ola Spjuth
Presentation at SLAS Europe 2019 in Barcelona on 28 june, 2019.
High-content microscopy in automated laboratories present many challenges for storing and processing data, and to build AI models to aid decision making. We have established an informatics system to serve a robotized cell profiling setup with incubators, liquid handling and high-content microscopy for microplates. The informatics system consists of computational infrastructure (CPUs, GPUs, storage), middleware (Kubernetes), imaging database and software (OMERO), and workflow system (Pachyderm) to perform online prioritization of new data, and automate the process from acquired images to continuously updated and deployed AI models. The AI methodologies include Deep Learning models trained on image data, and conventional machine learning models trained on data from Cell Painting experiments. The microservice architecture makes the system scalable and expandable, and a key objective is on improving screening and toxicity assessment using AI-aided intelligent experimental design.
The Advancement and Challenges in Computational Physics - PhdassistancePhD Assistance
For the last five decades, computational physics has been a valuable scientific instrument in physics. In comparison to using only theoretical and experimental approaches, it has enabled physicists to understand complex problems better. Computational physics was mostly a scientific activity at the time, with relatively few organised undergraduate study.
Ph.D. Assistance serves as an external mentor to brainstorm your idea and translate that into a research model. Hiring a mentor or tutor is common and therefore let your research committee know about the same. We do not offer any writing services without the involvement of the researcher.
Learn More: https://bit.ly/3AUvG0y
Contact Us:
Website: https://www.phdassistance.com/
UK NO: +44–1143520021
India No: +91–4448137070
WhatsApp No: +91 91769 66446
Email: info@phdassistance.com
On the large scale of studying dynamics with MEG: Lessons learned from the Hu...Robert Oostenveld
As part of the Human Connectome Project (HCP), which includes high-quality fMRI, anatomical MRI, DTi and genetic data from 1200 subjects, we have scanned and investigated a subset of 100 subjects (mostly comprised of pairs of twins) using MEG. The raw data acquired in the HCP has been analyzed using standard pipelines [ref1] and both raw and results at various levels of processing have been shared though the ConnectomeDB [ref2].
Throughout the process of the HCP we have not only analyzed (resting state) MEG data, but also have developed the data analysis protocols, the software and the strategies to achieve reproducible MEG connectivity results. The MEG data analysis software is based on FieldTrip, an open source toolbox [ref3], and is shared alongside the data to allow the analyses to be repeated on independent data.
In this presentation I will outline what the HCP MEG team has learned along the way and I will provide recommendations on what to do and what to avoid in making MEG studies on (resting state) connectivity more reproducible.
1. Larson-Prior LJ, Oostenveld R, Della Penna S, Michalareas G, Prior F, Babajani-Feremi A, Schoffelen JM, Marzetti L, de Pasquale F, Di Pompeo F, Stout J, Woolrich M, Luo Q, Bucholz R, Fries P, Pizzella V, Romani GL, Corbetta M, Snyder AZ; WU-Minn HCP Consortium. Adding dynamics to the Human Connectome Project with MEG. Neuroimage, 2013.
doi:10.1016/j.neuroimage.2013.05.056
2. Hodge MR, Horton W, Brown T, Herrick R, Olsen T, Hileman ME, McKay M, Archie KA, Cler E, Harms MP, Burgess GC, Glasser MF, Elam JS, Curtiss SW, Barch DM, Oostenveld R, Larson-Prior LJ, Ugurbil K, Van Essen DC, Marcus DS. ConnectomeDB-Sharing human brain connectivity data. Neuroimage, 2016. doi:10.1016/j.neuroimage.2015.04.046
3. Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Comput Intell Neurosci. 2011. doi:10.1155/2011/156869
Дмитрий Ветров. Математика больших данных: тензоры, нейросети, байесовский вы...Yandex
Лекция одного из самых известных в России специалистов по машинному обучению Дмитрия Ветрова, который руководит департаментом больших данных и информационного поиска на факультете компьютерных наук, работающим во ВШЭ при поддержке Яндекса.
These are the slides presented by Vladimir Litvak in the Open Science Panel discussion at the BIOMAG 2018 meeting in Philadelphia. See also https://www.frontiersin.org/research-topics/5158
The Advancement and Challenges in Computational Physics - PhdassistancePhD Assistance
For the last five decades, computational physics has been a valuable scientific instrument in physics. In comparison to using only theoretical and experimental approaches, it has enabled physicists to understand complex problems better. Computational physics was mostly a scientific activity at the time, with relatively few organised undergraduate study.
Ph.D. Assistance serves as an external mentor to brainstorm your idea and translate that into a research model. Hiring a mentor or tutor is common and therefore let your research committee know about the same. We do not offer any writing services without the involvement of the researcher.
Learn More: https://bit.ly/3AUvG0y
Contact Us:
Website: https://www.phdassistance.com/
UK NO: +44–1143520021
India No: +91–4448137070
WhatsApp No: +91 91769 66446
Email: info@phdassistance.com
On the large scale of studying dynamics with MEG: Lessons learned from the Hu...Robert Oostenveld
As part of the Human Connectome Project (HCP), which includes high-quality fMRI, anatomical MRI, DTi and genetic data from 1200 subjects, we have scanned and investigated a subset of 100 subjects (mostly comprised of pairs of twins) using MEG. The raw data acquired in the HCP has been analyzed using standard pipelines [ref1] and both raw and results at various levels of processing have been shared though the ConnectomeDB [ref2].
Throughout the process of the HCP we have not only analyzed (resting state) MEG data, but also have developed the data analysis protocols, the software and the strategies to achieve reproducible MEG connectivity results. The MEG data analysis software is based on FieldTrip, an open source toolbox [ref3], and is shared alongside the data to allow the analyses to be repeated on independent data.
In this presentation I will outline what the HCP MEG team has learned along the way and I will provide recommendations on what to do and what to avoid in making MEG studies on (resting state) connectivity more reproducible.
1. Larson-Prior LJ, Oostenveld R, Della Penna S, Michalareas G, Prior F, Babajani-Feremi A, Schoffelen JM, Marzetti L, de Pasquale F, Di Pompeo F, Stout J, Woolrich M, Luo Q, Bucholz R, Fries P, Pizzella V, Romani GL, Corbetta M, Snyder AZ; WU-Minn HCP Consortium. Adding dynamics to the Human Connectome Project with MEG. Neuroimage, 2013.
doi:10.1016/j.neuroimage.2013.05.056
2. Hodge MR, Horton W, Brown T, Herrick R, Olsen T, Hileman ME, McKay M, Archie KA, Cler E, Harms MP, Burgess GC, Glasser MF, Elam JS, Curtiss SW, Barch DM, Oostenveld R, Larson-Prior LJ, Ugurbil K, Van Essen DC, Marcus DS. ConnectomeDB-Sharing human brain connectivity data. Neuroimage, 2016. doi:10.1016/j.neuroimage.2015.04.046
3. Oostenveld R, Fries P, Maris E, Schoffelen JM. FieldTrip: Open Source Software for Advanced Analysis of MEG, EEG, and Invasive Electrophysiological Data. Comput Intell Neurosci. 2011. doi:10.1155/2011/156869
Дмитрий Ветров. Математика больших данных: тензоры, нейросети, байесовский вы...Yandex
Лекция одного из самых известных в России специалистов по машинному обучению Дмитрия Ветрова, который руководит департаментом больших данных и информационного поиска на факультете компьютерных наук, работающим во ВШЭ при поддержке Яндекса.
These are the slides presented by Vladimir Litvak in the Open Science Panel discussion at the BIOMAG 2018 meeting in Philadelphia. See also https://www.frontiersin.org/research-topics/5158
Thoughts on Knowledge Graphs & Deeper ProvenancePaul Groth
Thinking about the need for deeper provenance for knowledge graphs but also using knowledge graphs to enrich provenance. Presented at https://seminariomirianandres.unirioja.es/sw19/
Small Is Beautiful: Summarizing Scientific Workflows Using Semantic Annotat...Khalid Belhajjame
Scientific Workflows have become the workhorse of BigData analytics for scientists. As well as being repeatable and optimizable pipelines that bring together datasets and analysis tools, workflows make-up an important part of the provenance of data generated from their execution. By faithfully capturing all stages in the analysis, workflows play a critical part in building up the audit-trail (a.k.a. provenance) meta- data for derived datasets and contributes to the veracity of results. Provenance is essential for reporting results, reporting the method followed, and adapting to changes in the datasets or tools. These functions, however, are hampered by the complexity of workflows and consequently the complexity of data-trails generated from their instrumented execution. In this paper we propose the generation of workflow description summaries in order to tackle workflow complexity. We elaborate reduction primitives for summarizing workflows, and show how prim- itives, as building blocks, can be used in conjunction with semantic workflow annotations to encode different summariza- tion strategies. We report on the effectiveness of the method through experimental evaluation using real-world workflows from the Taverna system.
Bioinformatics databases: Current Trends and Future PerspectivesUniversity of Malaya
Data is the most powerful resource in any field or subject of study. In Biology, data comes from scientists and their actions, while any institution that makes sense of the data collected, will be in the forefront in their respective research field. In the beginning of any data collection endeavour, it is critical to find proper management techniques to store data and to maximise its utilisation. This presentation reflects upon the current trends and techniques of data modeling, architecture with a highlight on the uses of database, focusing on Bioinformatics examples and case studies. Finally, the future of bioinformatics databases is highlighted to give an overview of the modeling techniques to accommodate the biological data escalation in coming years.
Towards Automated AI-guided Drug Discovery LabsOla Spjuth
Presentation by Ola Spjuth (Uppsala University and Scaleout Systems) on 2019-10-16 at Swedish e-Science Academy 2019 in Lund, Sweden.
Research website at Uppsala University: https://pharmb.io
Scaleout Systems: https://scaleoutsystems.com
The interplay between data-driven and theory-driven methods for chemical scie...Ichigaku Takigawa
The 1st International Symposium on Human InformatiX
X-Dimensional Human Informatics and Biology
ATR, Kyoto, February 27-28, 2020
https://human-informatix.atr.jp
New learning technologies seem likely to transform much of science, as they are already doing for many areas of industry and society. We can expect these technologies to be used, for example, to obtain new insights from massive scientific data and to automate research processes. However, success in such endeavors will require new learning systems: scientific computing platforms, methods, and software that enable the large-scale application of learning technologies. These systems will need to enable learning from extremely large quantities of data; the management of large and complex data, models, and workflows; and the delivery of learning capabilities to many thousands of scientists. In this talk, I review these challenges and opportunities and describe systems that my colleagues and I are developing to enable the application of learning throughout the research process, from data acquisition to analysis.
Continuous modeling - automating model building on high-performance e-Infrast...
012517 ResumeJH Amex DS-ML
1. JEREMY HADIDJOJO
(510)-604-5316 hjeremy@umich.edu
4485C Randall Lab, 450 Church St, Ann Arbor, MI 48109-1040
PROFILE
• Computational physics, expertise in mathematical modeling, simulation and data analysis
• Extensive experience in:
scientific programming (MATLAB, Python, C++, parallel/GPU/HPC computing)
machine learning (SVM, neural networks, clustering, regression analysis, decision trees, PCA)
interdisciplinary collaboration (biologists, mathematicians, computer scientists, engineers)
scientific communication across disciplines and to non-scientific audiences
• Passion in research, coding, validating algorithms for machine learning and data science
• Strong analytical skills, able to derive and understand complex math behind algorithms/models
• Passion in exploring new technologies, especially in machine learning/data science
EDUCATION
University of Michigan, Ann Arbor August 2011 – present
Ph.D. in Physics, GPA 3.8/4.0, graduating May 2017
Nanyang Technological University, Singapore 2007 – 2011
B.Sc. in Physics with first-class Honours, Minor in Mathematics, GPA: 4.8/5.0
PROGRAMMING SKILLS & PROJECTS
Advanced: MATLAB, Mathematica, C/C++
Intermediate: OpenMP, Python (NumPy, SciPy, TensorFlow, SciKit-Learn, Panda, Cython)
Beginner: Theano, Embedded programming (Arduino, STM32F4)
1. Deep Learning of handwritted digits (MATLAB, Python) 2016
Coded from scratch object-oriented convolutional network in MATLAB, tested with MNIST hand-
written digit data. Reaches 99.4% accuracy with Python + Theano (GPU computing)
2. Large-scale cell mechanics simulation (C++) 2012 – present
Physical simulation of 2D tissue capable of handling 10,000+ cells. Written in object-oriented C++
with 25,000+ lines of code using (1) GSL for ODE integration, (2) BLAS/LAPACK for fast linear
algebra, and (3) OpenMP for parallel computation. MATLAB used for pre/post-processing and GUI.
3. High-performace timeseries analysis (MATLAB, Python, C++) 2013 – present
Developed highly-optimized codes for fast timeseries correlation. First version is MATLAB (parallel,
GPU), and second is Python calling compiled C++ routines (parallel OpenMP).
RESEARCH EXPERIENCE
New mechanism of planar cell chirality 2012 – present
• Devised a new framework of generating planar cell chirality through protein interaction
Developed mathematical model (pen & paper, Mathematica), performed numerical anal-
ysis (MATLAB) and simulation (C++ with BLAS/LAPACK, OpenMP)
Pattern formation of retinal cone photoreceptors 2012 – present
• Uncovered mechanisms that made patterns in zebrafish retina (published in PLoS ONE 2014)
Developed physical model based on experimental data, performed statistical analyses and
numerical simulation, and made prediction based on model.
Statistical analysis of noisy timeseries cell trajectories 2013 – presentt
• Searched for non-trivial correlation and causality between large timeseries of cell movement
2. Analyzed big data (terabytes), applied advanced statistical methods and machine
learning (SVM, clutering, mean-shift)
PUBLICATIONS
Hadidjojo J, Salbreux G, Lubensky DK (2017) Spontaneous Chiral Symmetry Breaking in Planar
Polarized Epithelia, Physical Review Letters (in preparation)
Nagashima M, Hadidjojo J, Barthel LK, Lubensky DK, Raymond PA (2017) Anisotropic Glial Scaf-
folding Shapes a Multiplex Photoreceptor Mosaic in Zebrafish Retina, eLife (submitted)
Raymond PA, Hadidjojo J, et al. (2014) Patterning the Cone Mosaic Array in Zebrafish Retina
Requires Specification of Ultraviolet-Sensitive Cones, PLoS ONE
Hadidjojo J, Cheong SA (2011) Equal Graph Partitioning on Estimated Infection Network as an
Effective Epidemic Mitigation Measure, PLoS ONE
CONFERENCES AND WORKSHOPS
Big Data Image Processing & Analysis Workshop (UC Irvine) 2016
Americal Physical Society (APS) March Meeting 2016 2016
Contributed talk: Planar Cell Chirality (PCC) from spontaneous symmetry breaking
EMBO Multi-level Modeling of Morphogenesis Workshop (John Innes Centre, UK) 2015
AWARDS & FELLOWSHIPS
Physics Department Graduate Fellowship August 2013
Awarded to 3 students based on past research and academic performance
Norman E. and Mary E. Barnett Graduate Fellowship January 2012
Awarded to 1 student in early PhD based on past research and academic performance
Physics Department Graduate Fellowship August 2011
Awarded to select incoming graduate students with outstanding undergraduate work
RELEVANT CLASSES
EECS 545: Statistical Machine Learning Fall 2015
Mathematics 630: Applied Stochastic Processes (audit) Fall 2014
Complex Systems 541: Nonlinear Dynamics Fall 2012
Complex Systems 510: Intro to Adaptive Systems Fall 2012
Physics 510: Statistical Mechanics Fall 2011
Complex Systems 535: Network Theory Fall 2011
HOBBIES AND OTHER ACTIVITIES
Photography, drone/quadcopter building and flying, electronics (Arduino), DIY in general.
Current spare-time project: tinkering with TensorFlow and Hadoop.