Successfully reported this slideshow.

Semantic Web for 360-degree Health: State-of-the-Art & Vision for Better Interoperability

3

Share

1 of 58
1 of 58

Semantic Web for 360-degree Health: State-of-the-Art & Vision for Better Interoperability

3

Share

Download to read offline

Ora Lassila and Amit Sheth, "Semantic Web for 360-degree Health: State-of-the-Art & Vision for Better Interoperability", Invited Talk at ONC-HHS Invitational Workshop on Next Generation Interoperability for Health, Washington DC, January 19-20, 2011.

Ora Lassila and Amit Sheth, "Semantic Web for 360-degree Health: State-of-the-Art & Vision for Better Interoperability", Invited Talk at ONC-HHS Invitational Workshop on Next Generation Interoperability for Health, Washington DC, January 19-20, 2011.

More Related Content

Related Books

Free with a 14 day trial from Scribd

See all

Related Audiobooks

Free with a 14 day trial from Scribd

See all

Semantic Web for 360-degree Health: State-of-the-Art & Vision for Better Interoperability

  1. 1. Ora Lassila Amit Sheth •  Principal Architect (Nokia •  LexisNexis Ohio Eminent Mobile Solutions); also an Scholar, Director, Ohio Center of advisor to Nokia’s top mgmt Excellence in Knowledge- •  Elected member of W3C’s enabled Computing (Kno.e.sis), Advisory Board since 1998 Wright State University •  Earlier: Research Fellow •  Educator, researcher, (Nokia Research), W3C Fellow entrepreneur – 2 companies, (MIT), Project Manager (CMU), products, deployed apps, W3C entrepreneur, etc. and biomedical community •  Ph.D from Helsinki University standards of Technology (CS) •  Earlier: UGA, Telcordia, Unisys, •  http://www.lassila.org/ Honeywell •  http://knoesis.org/amit
  2. 2. • Semantic Web Ora • some background • Semantic Web in use Amit • examples of applications in traditional clinical care to translational medicine • Challenges (and promise) Ora (technical) • what makes this difficult Amit (health) • why do we want to pursue it anyway
  3. 3. • Often characterized as the “next generation of the World Wide Web” •  Web content amenable to automation •  (current content intended for humans…)
  4. 4. • Often characterized as the “next generation of the World Wide Web” •  Web content amenable to automation •  (current content intended for humans…) • In reality, the Semantic Web is a vision of the future of (personal) computing •  machines working on behalf of their human users •  more autonomy, handling of unanticipated situations • Heavy reliance of knowledge representation & reasoning •  also multi-agent systems, other AI-based technologies
  5. 5. • At the core, the Semantic Web is about •  describing things (objects, concepts, services, …) •  querying the descriptions •  reasoning about the descriptions • As such, it is knowledge representation •  for the Web •  (or KR using standardized Web technologies) • (in comparison, the “old Web” was really about documents and finding them…)
  6. 6. • Motivated by the need for automation •  automation requires interoperability (via standards) •  heavy process, high up-front investment •  (alternative: hand-crafted but “brittle” programs…) • Interoperability achieved by exposing meaning •  accessible semantics •  note: interoperability of any two systems can be achieved via engineering, but this does not scale • Automation → autonomy •  prevailing paradigm: agent-based systems •  implies reasoning, planning, interoperable representations of knowledge
  7. 7. • Contrary to “Web 2.0”, Semantic Web aims at achieving many things “ad hoc” •  e.g., ad hoc mash-ups by non-computer savvy people • Shared (and accessible) semantics is the key to interoperability • Semantic Web introduces a fundamentally different approach to standardization •  standardize how to say things and not what to say •  ontological techniques allow “delayed semantic commitment”
  8. 8. • Semantic Web is built in a layered manner • Not everybody needs all the layers … Queries: SPARQL, Rules: RIF Semantic Web Rich ontologies: OWL Simple data models & taxonomies: RDF Schema Uniform metamodel: RDF + URI Encoding structure: XML Encoding characters : Unicode
  9. 9. • Achieve for data what Web did to documents • Relationship with the original Semantic Web vision: no AI, no agents, no autonomy • Interoperability is still very important •  interoperability of formats •  interoperability of semantics • Enables interchange of large data sets •  (thus very useful in, say, collaborative research) • Semantic Web vision is largely predicated on the availability of data •  Linked Data is a movement that gets us there
  10. 10. Tech assimilated in life Web of Sensors, Devices/IoT Situations, - 40 billion sensors, 5 billion mobile connections 2007 Events Web 3.0 Objects Web of people - social networks, user-created casual content Patterns Web of resources Web 2.0 - data, service, data, mashups Keywords Web of databases 1997 - dynamically generated pages - web query interfaces Web of pages - text, manually created links Web 1.0 - extensive navigation
  11. 11. ...needs a connection Hypothesis Validation Experiment design Predictions Personalized medicine Biomedical Informatics Etiology Genome More advanced capabilities for Pathogenesis Transcriptome Clinical findings search, Proteome Diagnosis Genbank Metabolome integration, Pubmed Prognosis Physiome analysis, Treatment ...ome linking to new insights Uniprot Clinical and discoveries! Trials.gov Medical Informatics Bioinformatics
  12. 12. text User-contributed Scientific Health NCBI Content (Informal) Clinical Data Laboratory Literature Information Public Datasets Experts: Data Services GeneRifs WikiGene PubMed Elsevier Genome, Lab tests, 300 Documents Consumer: Protein DBs Personal iConsult RTPCR, Published Online Blogs new sequences health history daily Mass spec each day Social Networks Search, browsing, complex query, integration, workflow, analysis, hypothesis validation, decision support.
  13. 13. • W3C Semantic Web Health Care & Life Sciences Interest Group: http://www.w3.org/2001/sw/hcls/ • Clinical Observations Interoperability: EMR + Clinical Trials: http://esw.w3.org/HCLS/ ClinicalObservationsInteroperability • National Center for Biomedical Ontologies: http://bioportal.bioontology.org/
  14. 14. • Status: In use continuously since 01/2006 • Where: Athens Heart Center & its partners and labs • What: Use of semantic Web technologies for clinical decision support
  15. 15. Examples demonstrating use of Semantic Web for Health Care and Life Sciences research projects and operational clinical or research applications
  16. 16. Details: http://knoesis.org/library/resource.php?id=00004
  17. 17. Annotate ICD9s Annotate Doctors Lexical Annotation Insurance Formulary Level 3 Drug Interaction Demo at: http://knoesis.org/library/demos/ Drug Allergy
  18. 18. formulary_ non_drug_ interaction_ property formulary reactant property indication indication_ property owl:thing monograph property _ix_class prescription interaction_ _drug_ with_non_ brandname_ prescription brand_name drug_reactant prescription individual _drug interaction _drug_ property brandname_ brandname_ composite prescription interaction_ undeclared _drug_ with_mono interaction_ generic graph_ix_cl with_prescri cpnum_ generic_ ass ption_drug group composite generic_ individual
  19. 19. • Status: Completed research • Where: NIH • What: queries across integrated data sources •  Enriching data with ontologies for integration, querying, and automation •  Ontologies beyond vocabularies: the power of relationships
  20. 20. Gene name Glycosyltransferase Interactions GO gene Sequence PubMed OMIM Congenital muscular dystrophy Adapted from: Olivier Bodenreider, presentation at HCLS Workshop, WWW07 http://knoesis.org/library/resource.php?id=00014
  21. 21. (GeneID: 9215) has_associated_disease Congenital muscular dystrophy, type 1D has_molecular_function Acetylglucosaminyl- transferase activity Adapted from: Olivier Bodenreider, presentation at HCLS Workshop, WWW07
  22. 22. SELECT DISTINCT ?t ?g ?d { ?t is_a GO:0016757 . glycosyltransferase ?g has molecular function ?t . GO:0016757 ?g has_associated_phenotype ?b2 . ?b2 has_textual_description ?d . isa FILTER (?d, “muscular distrophy”, “i”) . GO:0008194“congenital”,GO:0016758 FILTER (?d, “i”) } acetylglucosaminyl- GO:0008375 transferase has_molecular_function acetylglucosaminyl- GO:0008375 transferase LARGE EG:9215 Muscular dystrophy, MIM:608840 has_associated_phenotype congenital, type 1D From medinfo paper. Adapted from: Olivier Bodenreider, presentation at HCLS Workshop, WWW07
  23. 23. • Status: Completed research • Where: NIH • What: Understanding the genetic basis of nicotine dependence. Integrate gene and pathway information and show how three complex biological queries can be answered by the integrated knowledge base. • How: Semantic Web technologies (especially RDF, OWL, and SPARQL) support information integration and make it easy to create semantic mashups (semantically integrated resources).
  24. 24. • NIDA study on nicotine dependency • List of candidate genes in humans • Analysis objectives include: o Find interactions between genes o Identification of active genes – maximum number of pathways o Identification of genes based on anatomical locations • Requires integration of genome and biological pathway information
  25. 25. Genome and pathway information integration KEGG Reactome • pathway • pathway • protein HumanCyc • protein • pmid • pathway • pmid • protein • pmid Entrez Gene • GO ID • HomoloGene ID GeneOntology HomoloGene http://knoesis.org/library/resource.php?id=00221
  26. 26. Entrez Knowledge Model (EKoM) BioPAX ontology
  27. 27. • Status: Research prototype – in regular lab use • Where: Center for Tropical and Emerging Global Diseases (CTEGD), UGA • What: Semantics and Services Enabled Problem Solving Environment for Trypanosoma cruzi • Who: Kno.e.sis, UGA, NCBO
  28. 28. Ohio Center of Excellence in Knowledge-enabled Computing (Kno.e.sis), Wright State University Tarleton Research Group, Center for Tropical and Emerging Global Diseases(CTEGD), University of Georgia Large Scale Distributed Information Systems (LSDIS). University of Georgia National Center for Biological Ontologies (NCBO), Stanford University The Wellcome Trust Sanger Institute, Cambridge, UK The Oswaldo Cruz Institute (Fiocruz), Brazil
  29. 29. • T. cruzi is a protozoan parasite that causes Chagas Disease or American trypanosomiasis • Chagas disease is the leading cause of death in Latin America where around 18 million people are infected with this parasite T. Brucei surrounded by red blood cells in a smear of infected blood. • Related parasites include, (Copyright: Jürgen Berger and Dr. Peter Overath, Max Planck Institute for Trypanosoma brucei and Developmental Biology, Tübengen) Leishmania major that causes African trypanosomiasis and leishmaniasis, respectively.
  30. 30. Trykipedia - a Wiki-based platform for collaboration of Parasite Research Community
  31. 31. •  Data Resources  Internal lab data (from Tarleton Research Group)  Gene Knockout, Strain Creation, Microarray, and Proteome  External databases (TriTrypDB, ProtozoaDB, Drug Bank, etc. ) •  Ontologies  Parasite Lifecycle Ontology (PLO)  Parasite Experiment Ontology (PEO) •  PKR supports complex biological queries related to T.cruzi drugs, vaccination, or gene knockout targets; for example,  Find all genes with proteomic expression in mammalian lifecycle stage with GPI anchor or signal peptide predictions.  Find genes annotated as potential vaccine candidates.  Find all genes with proteomic expression evidence in the mammalian host lifecycle stages for T. cruzi
  32. 32. Gene Name Sequence Extraction Gene Knockout and Strain Creation* Related Queries from Biologists Drug 3‘ & 5’ Resistant Region Plasmid Gene Name Plasmid Construction •  List all groups in the lab that used T.Cruzi Knockout Construct a Target Region Plasmid? sample Plasmid Transfection •  Which ?researcher created a new strain of the parasite (with ID = Transfecte d Sample 66)? •  An experiment was not successful Drug Selection Cloned Sample Selected Sample – has this experiment been Cell Cloning conducted earlier? What were the Cloned results? Sample *T.cruzi Semantic Problem Solving Environment Project, Courtesy of D.B. Weatherly and Flora Logan, Tarleton Lab, University of Georgia
  33. 33. Complex queries can also include: - on-the-fly Web services execution to retrieve additional data -  inference rules to make implicit knowledge explicit
  34. 34. 1.  Describe drug user’s knowledge, attitudes, and behaviors related to illicit use of OxyContin® 2.  Describe temporal patterns of non-medical use of OxyContin® tablets as discussed on Web-based forums 3.  Collaboration between Kno.e.sis and CITAR (Center for Interventions, Treatment and Addictions Research) at Wright State Univ.
  35. 35. • Volatile nature of execution environments •  May have an impact on multiple activities/ tasks in the workflow • HF Pathway •  New information about diseases, drugs becomes available •  Affects treatment plans, drug-drug interactions • Need to incorporate the new knowledge into execution •  capture the constraints and relationships between different tasks activities
  36. 36. New knowledge about treatment found during the execution of the pathway New knowledge about drugs, drug drug interactions
  37. 37. Diabetes mellitus adversely affects the outcomes in patients with myocardial infarction (MI), due in part to the exacerbation of left ventricular (LV) remodeling. Although angiotensin II type 1 receptor blocker (ARB) has been demonstrated to be effective in the treatment of heart failure, information about the potential benefits of ARB on advanced LV failure associated with diabetes is lacking. To induce diabetes, male mice were injected intraperitoneally with streptozotocin (200 mg/kg). At 2 weeks, anterior MI was created by ligating the left coronary artery. These animals received treatment with olmesartan (0.1 mg/kg/day; n = 50) or vehicle (n = 51) for 4 weeks. Diabetes worsened the survival and exaggerated echocardiographic LV dilatation and dysfunction in MI. Treatment of diabetic MI mice with olmesartan significantly improved the survival rate (42% versus 27%, P < 0.05) without affecting blood glucose, arterial blood pressure, or infarct size. It also attenuated LV dysfunction in diabetic MI. Likewise, olmesartan attenuated myocyte hypertrophy, interstitial fibrosis, and the number of apoptotic cells in the noninfarcted LV from diabetic MI. Post-MI LV remodeling and failure in diabetes were ameliorated by ARB, providing further evidence that angiotensin II plays a pivotal role in the exacerbated heart failure after diabetic MI. possibly ARB plays role in heart failure Angiotensin II type 1 receptor blocker attenuates exacerbated left ventricular remodeling and failure in diabetes-associated myocardial infarction., Matsusaka H, et. al.
  38. 38. Disease possibly plays role in Angiotension Receptor Blocker (ARB) Ontology: A Framework for Schema-Driven Relationship Discovery from Unstructured Text, Ramakrishnan, et. al., ISWC 2006, LNCS 4273, pp. 583-596
  39. 39. •  Matching medical requirements with availability of medical resources (Mumbai, India) •  Project HERO Helpline for Emergency Response Operations •  For patients seeking for immediate medical help •  Medical awareness in rural India •  mMitra, info. service during pregnancy and childhood emergency Medical Medical Information Emergency Resourc bridge es
  40. 40. • Any specific problem (typically) has a specific solution that does not require Semantic Web technologies • Q: Why then is the Semantic Web attractive? A: For future-proofing Semantic Web can be a solution to those problems and situations that we are yet to define
  41. 41. • Cultural resistance (“this smacks of AI…”) • Unfamiliar technology (e.g., reasoning) • Often implies complex representational models •  procedural programs vs. declarative data • Unclear business models • Also, actual technical challenges •  scalability of query processing •  complexity (and thus scalability) of reasoning •  scalability of access control •  …
  42. 42. • (merely an observation of what you may encounter…) Source: Mindlab, U of Maryland • What makes Semantic Web attractive and worth pursuing is…
  43. 43. an Dictionary) (Source: Oxford Americ • Serendipity in interoperability •  can we interoperate with systems, devices and/or services we knew nothing about at design time? • Serendipity in information reuse •  with accessible semantics, this becomes easier… • Serendipity in information integration •  can information from independent sources be combined? •  even simple forms of reasoning can help
  44. 44. • Semantic Web was designed to •  accommodate different points of view •  be flexible about what it can express (not preferential towards any particular domain or application) • Combining information in new ways •  we cannot anticipate all the possible ways in which information is used, combined ⇒  there is value to merely making information (data) available •  using Semantic Web technologies lowers the threshold for “serendipitous reuse”
  45. 45. Insurance, Clinical Care Financial Aspects Follow up, Lifestyle Genetic Tests… Profiles Social Media Clinical Trials
  46. 46. NIH FDA CDC (Research) Universities, Pharmaceutical AMCs Companies Patients, Public CROs Hospitals Doctors Payors From FDA, CDC Translation 1: Genomic Research and Clinical Practice Translation 2: Clinical Research and Clinical Practice Slide by: Vipul Kashyap
  47. 47. • For each component in 360-degree health care, we have data, processes, knowledge and experience. Interoperability solutions need to encompass all these! •  Possibly largest growth in data will be in sensors (eg Body Area Networks, Biosensors) and social content. Extensive use of mobile phones. Credit: ece.virginia.edu
  48. 48. • Semantic Web is an “interoperability technology” • Linked Data is a step in the right direction • Many examples of viable usage of Semantic Web technologies • Words of warning about deployment • For health, Semantic Web provides the needed interoperability, and can accommodate all necessary “points of view” • Significant research challenges remain as Health presents the most complex domain
  49. 49. • Researchers: Satya Sahoo, Dr. Priti Parikh, Pablo Mendes, Cartic Ramakrishnan, and Kno.e.sis team • Collaborators: Athens Heart Center (Dr. Agrawal), NLM (Olivier Bodenreider), CCRC- UGA (Will York), UGA (Tarleton), Bioinformatics-WSU (Raymer) • Funding: NIH/NCRR, NIH/NLBHI (R01), NSF http://knoesis.org
  50. 50. 1.  A. Sheth, S. Agrawal, J. Lathem, N. Oldham, H. Wingate, P. Yadav, and K. Gallagher, Active Semantic Electronic Medical Record, Intl Semantic Web Conference, 2006. 2.  Satya Sahoo, Olivier Bodenreider, Kelly Zeng, and Amit Sheth, An Experiment in Integrating Large Biomedical Knowledge Resources with RDF: Application to Associating Genotype and Phenotype Information WWW2007 HCLS Workshop, May 2007. 3.  Satya S. Sahoo, Kelly Zeng, Olivier Bodenreider, and Amit Sheth, From "Glycosyltransferase to Congenital Muscular Dystrophy: Integrating Knowledge from NCBI Entrez Gene and the Gene Ontology, Amsterdam: IOS, August 2007, PMID: 17911917, pp. 1260-4 4.  Satya S. Sahoo, Olivier Bodenreider, Joni L. Rutter, Karen J. Skinner , Amit P. Sheth, An ontology-driven semantic mash-up of gene and biological pathway information: Application to the domain of nicotine dependence, Journal of Biomedical Informatics, 2008. 5.  Cartic Ramakrishnan, Krzysztof J. Kochut, and Amit Sheth, " A Framework for Schema-Driven Relationship Discovery from Unstructured Text", Intl Semantic Web Conference, 2006, pp. 583-596 6.  Satya S. Sahoo, Christopher Thomas, Amit Sheth, William S. York, and Samir Tartir, " Knowledge Modeling and Its Application in Life Sciences: A Tale of Two Ontologies", 15th International World Wide Web Conference (WWW2006), Edinburgh, Scotland, May 23-26, 2006. 7.  Satya S. Sahoo, Olivier Bodenreider, Pascal Hitzler, Amit Sheth and Krishnaprasad Thirunarayan, ' Provenance Context Entity (PaCE): Scalable provenance tracking for scientific RDF data.’ SSDBM, Heidelberg, Germany 2010. •  Papers: http://knoesis.org/library •  Demos at: http://knoesis.wright.edu/library/demos/

×