Downloaded 20 times
Uploaded byMichel Dumontier
2011 ebi industry workshop
This document describes how semantic web technologies can be used to predict druglikeness and toxicity by integrating data and services. It discusses using ontologies like SIO and CHEMINF to formally define concepts like drug-likeness. Services built using SADI consume SMILES strings and annotate molecules with descriptors to determine if they satisfy definitions. Multiple chemical services were created to analyze caffeine and determine it has drug-like properties based on the Lipinski rule of five. The semantic web allows disparate data sources like ChEBI and Bio2RDF to be integrated and queried as a single knowledge base.
More Related Content
Similar to 2011 ebi industry workshop
More from Michel Dumontier
2011 ebi industry workshop
- 1. Predicting Druglikeness andToxicity from Integrated Data and Services on the Life Science Semantic Web1Michel Dumontier, Ph.D.Associate Professor of Bioinformatics, Department of Biology, School of Computer Science, Institute of Biochemistry, Carleton UniversityProfesseur AssociĂŠ, DĂŠpartement dâinformatique et de gĂŠnielogiciel, UniversitĂŠ LavalOttawa Institute of Systems BiologyOttawa-Carleton Institute of Biomedical Engineering2011-EBI-Industry-SW::Dumontier
- 2. Is caffeine adrug-like molecule?Is acetaminophen toxic?22011-EBI-Industry-SW::Dumontier
- 3. Finding the rightinformation to answer a question is hardand sometimes requires a sophisticated workflow32011-EBI-Industry-SW::Dumontier
- 4.
- 5. What if wecould answer a question by automatically building a knowledge base using both data and services?52011-EBI-Industry-SW::Dumontier
- 6. The Semantic Webis a web of knowledge.6It is about standards for publishing, sharing and querying knowledge drawn from diverse sourcesIt enables the answering of sophisticated questions2011-EBI-Industry-SW::Dumontier
- 7. Is caffeine adrug-like molecule?To answer this question we need to know: what âdrug like moleculeâ really means
- 8. caffeineâs molecularstructure
- 9. the abilityto compute the relevant attributes
- 10. determine whethercaffeine satisfies the requirements of being âdrug likeâ 72011-EBI-Industry-SW::Dumontier
- 11. Lipinski Rule ofFiveRule of thumb for druglikeness (orally active in humans) (4 rules with multiples of 5)mass of less than 500 Daltonsfewer than 5 hydrogen bond donorsfewer than 10 hydrogen bond acceptorsA partition coefficient value between -5 and 5We need a more formal (machine understandable) description of a âdrug-like moleculeâ which specifies values for chemical descriptors82011-EBI-Industry-SW::Dumontier
- 12. ontology as astrategy to formally represent knowledge92011-EBI-Industry-SW::Dumontier
- 13. The Web OntologyLanguage (OWL) Has Explicit SemanticsCan therefore be used to capture knowledge in a machine understandable way102011-EBI-Industry-SW::Dumontier
- 14. Semanticscience Integrated Ontology(SIO)OWL2 ontology900+ classes covering basic types (physical, processual, abstract, informational) with an emphasis on biological entities169 basic relations (mereological, participatory, attribute/quality, spatial, temporal and representational)axioms can be used by reasoners to generate inferences for consistency checking, classification and answering questions about life science knowledgeembodies emerging ontology design patterns specifies the representation of knowledgedereferenceable URIssearchable in the NCBO bioportalAvailable at http://semanticscience.org/ontology/sio.owl112011-EBI-Industry-SW::Dumontier
- 15.
- 16. The Chemical InformationOntology (CHEMINF)100+ chemical descriptors50+ chemical qualitiesRelates descriptors to their specifications, the software that generated them (along with the running parameters, and the algorithms that they implement)Contributors: Nico Adams, Leonid Chepelev, Michel Dumontier, Janna Hastings, EgonWillighagen, Peter Murray-Rust, Cristoph Steinbeck13http://semanticchemistry.googlecode.com2011-EBI-Industry-SW::Dumontier
- 17. Molecular structure canbe represented using a SMILES string, which is a common representation of the chemical graph14Cn1cnc2n(C)c(=O)n(C)c(=O)c12ball & stick model for caffeineSMILES string for caffeine2011-EBI-Industry-SW::Dumontier
- 18. Lipinski Rule ofFiveEmpirically derived ruleset for druglikeness (4 rules with multiples of 5)mass of less than 500 Daltonsfewer than 5 hydrogen bond donorsfewer than 10 hydrogen bond acceptorsA partition coefficient value between -5 and 5A formal description using OWL:152011-EBI-Industry-SW::Dumontier
- 19. What we thenneed are services that will consume SMILES strings and annotate the molecule with the required chemical descriptors 16then we can reason about whether it satisfies the drug-likeness definition2011-EBI-Industry-SW::Dumontier
- 20. Semantic Automated Discoveryand Integrationhttp://sadiframework.orgSADI is a framework to create Semantic Web services using OWL classes as service inputs and outputsMark Wilkinson, UBCMichel Dumontier, Carleton UniversityChristopher Baker, UNB172011-EBI-Industry-SW::Dumontier
- 21. Create code stubsusing the ontologyPublish the ontology to a web-accessible locationhttp://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl Make sure that the class names are resolvable(easy when using the hash notation)http://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#smiles-moleculehttp://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#logp-moleculehttp://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#hbdc-moleculehttp://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#hdba-moleculehttp://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#lipinksi-druglike-moleculeDownload/checkout the codehttp://sadiframework.orgRun the code generator (Java, Perl, python)specify the URIs that correspond to input and output typesImplement the functionalityWe used the Chemistry Development Kit (CDK) to implement 4 services182011-EBI-Industry-SW::Dumontier
- 22. Responds to aGET operation by providing the service description in RDF conforms to Feta (BioMoby, myGrid)19curl http://cbrass.biordf.net/logpdc/logpc<rdf:RDFxmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:j.0="http://www.mygrid.org.uk/mygrid-moby-service#" > <rdf:Descriptionrdf:about=""> <j.0:hasServiceDescriptionText>no description</j.0:hasServiceDescriptionText> <j.0:hasServiceNameText rdf:datatype="http://www.w3.org/2001/XMLSchema#string">logpc</j.0:hasServiceNameText> <j.0:hasOperation rdf:resource="#operation"/> <rdf:typerdf:resource="http://www.mygrid.org.uk/mygrid-moby-service#serviceDescription"/> </rdf:Description> <rdf:Descriptionrdf:about="#input"> <j.0:objectType rdf:resource="http://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#smilesmolecule"/> <rdf:typerdf:resource="http://www.mygrid.org.uk/mygrid-moby-service#parameter"/> </rdf:Description> <rdf:Descriptionrdf:about="#operation"> <j.0:outputParameter rdf:resource="#output"/> <j.0:inputParameter rdf:resource="#input"/> <rdf:typerdf:resource="http://www.mygrid.org.uk/mygrid-moby-service#operation"/> </rdf:Description> <rdf:Descriptionrdf:about="#output"> <j.0:objectType rdf:resource="http://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#alogpsmilesmolecule"/> <rdf:typerdf:resource="http://www.mygrid.org.uk/mygrid-moby-service#parameter"/> </rdf:Description></rdf:RDF>2011-EBI-Industry-SW::Dumontier
- 23. Responds to aPOST containing service input with a service output in RDF20The query is in RDF:<rdf:RDFxmlns="http://semanticscience.org/sadi/ontology/caffeine.rdf#"xmlns:so="http://semanticscience.org/sadi/ontology/lipinskiserviceontology.owl#"xmlns:owl="http://www.w3.org/2002/07/owl#"xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"xmlns:sio="http://semanticscience.org/resource/"xmlns:xsd="http://www.w3.org/2001/XMLSchema#"> <so:smilesmoleculerdf:about="http://semanticscience.org/sadi/ontology/caffeine.rdf#m"> <sio:SIO_000008 rdf:resource = "http://semanticscience.org/sadi/ontology/caffeine.rdf#msmiles"/> </so:smilesmolecule> <sio:CHEMINF_000018 rdf:about = "http://semanticscience.org/sadi/ontology/caffeine.rdf#msmiles"> <sio:SIO_000300 rdf:datatype="xsd:string">Cn1cnc2n(C)c(=O)n(C)c(=O)c12</sio:SIO_000300> </sio:CHEMINF_000018></rdf:RDF>The response is in RDF: <rdf:Descriptionrdf:about="http://semanticscience.org/sadi/ontology/caffeine.rdf#mdalogp"> <rdf:typerdf:resource="http://semanticscience.org/resource/CHEMINF_000251"/> <j.0:SIO_000300 rdf:datatype="http://www.w3.org/2001/XMLSchema#double">-0.4311000000000006</j.0:SIO_000300> </rdf:Description>2011-EBI-Industry-SW::Dumontier
- 24. 61 Chemical SemanticWeb Servicesthese and an increasing number of semantic web services are registered at http://sadiframework.org/registry/services/212011-EBI-Industry-SW::Dumontier
- 25.
- 26. 23Semantic Health andResearch EnvironmentSHARE is an application that execute (SPARQL) queries as workflows over SADI Services2011-EBI-Industry-SW::Dumontier
- 27. âReckoningâdynamic discovery ofinstances of OWL classes through synthesis and invocation of a Web Service workflow capable of generating data described by the OWL class restrictions, followed by reasoning to classify the data into that ontology242011-EBI-Industry-SW::Dumontier
- 28. ChEBI publishes (non-SW)data!252011-EBI-Industry-SW::Dumontier
- 29. Bio2RDF provides ChEBIin RDF ď262011-EBI-Industry-SW::Dumontier
- 30. Bio2RDF covers themajor biological databases272011-EBI-Industry-SW::Dumontier
- 31. Bio2RDFâs RDFized datafits together28
- 32. Resource Description Framework(RDF)Allows one to talk about anythingUniform Resource Identifier (URI) can be used as entity names Bio2RDF specifies the naming conventionhttp://bio2rdf.org/uniprot:P05067 is a name for Amyloid precursor proteinhttp://bio2rdf.org/omim:104300 is a name for Alzheimer diseaseuniprot:P05067omim:10430029
- 33. Life Science DatasetRegistry Coordinates NamingProvides stable URI patterns for records and the entities they describe.Directory Service~1500 datasets & dozens of resolvers.Discovery ServiceRegistry links entities to records and their representations (RDF/XML, HTML, etc) and provider (Bio2RDF, Uniprot)Redirection ServiceAutomatic redirection to data provider document30Stanford : 22-04-2010
- 34. Bio2RDF is nowserving over 40 billion triples of linked biological data312011-EBI-Industry-SW::Dumontier
- 35. Bio2RDF is aframework to create and provision linked data networks32Francois Belleau, Laval UniversityMarc-Alexandre Nolin, Laval UniversityPeter Ansell, Queensland University of TechnologyMichel Dumontier, Carleton University
- 36. Bio2RDF is partof a growing web of linked data33âLinking Open Data clouddiagram, by Richard Cyganiak and Anja Jentzsch. http://lod-cloud.net/â 2011-EBI-Industry-SW::Dumontier
- 37. something you canlookup or search for with rich descriptions342011-EBI-Industry-SW::Dumontier
- 38. 35SPARQL is thenewcool kid on the query blockSQLSPARQL2011-EBI-Industry-SW::Dumontier
- 39. Query for logp362011-EBI-Industry-SW::Dumontier
- 40.
- 41. Query: Is caffeinea drug-like molecule?382011-EBI-Industry-SW::Dumontier
- 42.
- 43. BenefitsData remains distributedâ as the internet was meant to be!Data is not âexposedâ as a SPARQL endpointgreater provider-control over computational resourcesService invocation is straightforward and matchmaking by reasoning about ontology-based input/output descriptions402011-EBI-Industry-SW::Dumontier
- 44. Is acetaminophen toxic?Classicalapproaches involve decision trees or machine learning over validated data.Algorithms are often proprietary, even by the regulatory agenciesIssues around which data was used, and what the informative parameters are, and how easily can new information affect the outcomes?412011-EBI-Industry-SW::Dumontier
- 45. OWLED2011 : Large-ScaleBoolean Feature Based Trees as OWL ontologies422011-EBI-Industry-SW::Dumontier
- 46. DL Reasoners giveExplanations432011-EBI-Industry-SW::Dumontier
- 47. SummarySemantic Web technologiesoffer tantalizing ability to create and share data and services for drug discoveryBio2RDF provides linked life science dataSADI provides a framework to provide semantic web servicesSHARE allows us to simultaneously query and reason about data and services represented using RDF/OWLExpressive ontologies can be used to make toxicity decisions transparent442011-EBI-Industry-SW::Dumontier
- 48. 45AcknowledgementsCHEMINF GroupLeo ChepelevJannaHastingsEgonWillighagenNico AdamsBio2RDF: Peter Ansell, Francois Belleau, Allison Callahan, Jacques Corbeil, Jose Cruz-Toledo, Alex De Leon, Steve Etlinger, James Hogan, Nichealla Keath, Jean Morissette, Marc-Alexandre Nolin, Nicole Tourigny, Philippe Rigault and, Â Paul Roe SADI: Christopher Baker, Melanie Courtot, Jose Cruz-Toledo, Steve Etlinger, Nichealla Keath, Artjom Klein, Luke McCarthy, Silvane Paixao, Ben Vandervalk, Natalia Villanueva-Rosales, Mark WilkinsonToxicity GroupLeo ChepelevDana Klassen2011-EBI-Industry-SW::Dumontier
- 49.
Editor's Notes
- #4Â Canât answer questions that require background knowledge