State of the GOC Rama Balakrishnan - PowerPoint Presentation

1. State of the GOCRama BalakrishnanGenetics DepartmentTechnology and Innovation Park

2. About GOCStarted as a joint project of 3 MODs in 1998SGD, MGI and FlybaseMakes us one of the founding membersGoal is to provide a common controlled vocabulary for describing genes and gene products in all organismsA way to capture biological knowledge in a written and computable form

3. Role of CherryLab We contribute to the ontology developmentWe annotate yeast genes (gene products) to GO termsWe host the production servers for GOC and AmiGO

4. Reactome

5. Three aspects of GO ProjectOntologiesDevelopment and maintanence of the ontologies AnnotationsAssociating the GO terms with gene products Tools to create and analyze data

6. Ontologies: The Scope of GO1. Molecular Function (how)e.g. protein kinase activity2. Biological Process (what)e.g. cell cycle3. Cellular Component (where)e.g. mitochondrionGO terms aim to describe the ‘normal’ functions/ processes/locations that gene products are involved in

7. Anatomy of a GO term10/8/14a.b.c.

8. GO AnnotationConnections or associations between gene products and GO terms 1. Gene or gene product identifiere.g. Q9ARH1 2. GO term IDe.g. GO:0004674 (protein serine/threonine kinase)3. Reference ID e.g. PubMed ID: 12374299 GO_REF:00000014. Evidence codee.g. IDA..and also in some cases: Qualifiers available to modify interpretation of annotationNOTcontributes_tocolocalizes_with ‘With’ column information, to provide further information on the method (evidence code)

9. AnnotationsOntology Ontology and Annotation Development MOD curators contribute to most of the Terms and Annotations In the recent past several groups have contributed terms and annotations

10. Ontology development 40K termsTerms are requested by curators, communityOntology editors add most of the termsTermgenie can also “grant” term requestsAligning the ontologies with CheBI, SAO, RHEAExpanding the ontologies with several external groups (Syscilia consortium, Giardia) Cell cycle overhaulBiological phase is now its own termTerms to annotate viral and host genes

11. Annotations53 million genes (products) from over 400,000 species have an annotation4 million manual annotations Annotations are submited in a standard 17 column gene_associations file (aka GAF)Capturing more specificity for annotations (aka col-16)Several new annotation groups are contributing annotations

12. GO Annotation RepositoryAmiGOValidationAnd more…GAF file

13. How are these annotations consumed? Annotations are the product of the GOCBench biologists just want to know the function of the genes of their interestOmics people use term enrichment or slim mapping to find patterns in their gene lists

14. How does the GOC bring its data to the users?WebsiteAmiGODownloadable filesTerm enrichment tool

15. New website

16. AmiGOGOC’s web application for search and retrieval of Ontology terms and annotationsMoved away from MySQL to Solr searchSupports Term enrichment and slim mappingAlso provides an online sql querying interface

17. Coming soon… LEGO modelsAbility to annotate complexes as objects

18. ExamplePMID: 7939631“ In medium depleted of phosphate, Pho81 is bound to and inhibits the kinase activity of PHO81-PHO85; inhibition of PHO80-PHO85 allows underphosphorylated PHO4 to activate transcription of PHO5.

19. GO Annotation for PHO81 with Annotation ExtensionGene NameGO TermReferenceEvidenceAnnotation ExtensionPHO81GO:4861, cyclin-dependent protein serine/threonine kinase inhibitor activityPMID: 7939631Inferred from direct assayhas_direct_input: PHO80-PHO85 complex (GO:0000307)part_of: cellular response to phosphate starvation (GO:0016036)

20. This is still not satisfying! Lot more data/story in the paper!How can we represent the whole story?Pathway like representation of GO annotationsLEGO model (Logical Extension of GO)“ In medium depleted of phosphate, Pho81 is bound to and inhibits the kinase activity of PHO81-PHO85; inhibition of PHO80-PHO85 allows underphosphorylated PHO4 to activate transcription of PHO5.”

21. LEGOCore unitA gene product (GP) carrying out a MF at a particular location/componentConnections to other units:What other brick(s) it is connected to (i.e. its “targets”)What larger structure(s) the brick is used to build (i.e. biological process(es))

22. LEGOCurrently, GPs have separate MF, CC, BP annotationsan MF annotation states that a particular GP executes a particular MF in some CC as part of some BP (incomplete)In LEGO, a particular GP executes a particular MF in a particular CC as part of a particular BPLEGO is backwards compatible with current annotations

23. LEGOCurrently, the causal relations integral to the study of molecular biology can be represented only very generallyE.g. GP involved_in regulation of MAP kinase activityIn LEGO, we can also representWhich gene product is being regulatedWhat is regulated (the amount of gene product, the activity or stability, the location)Whether the regulation is direct or indirectWhat larger processes this is part ofIn LEGO, we can represent pathways and coordinated processes, and build these representations from individual annotations from primary literature publications

24. Pho81 pathway in LEGOPho80-Pho85 complexPho81Pho4pPHO5Pho5p

25. Killer App Term enrichmentTool used by most scientists to identify shared significant terms for a gene setBuzz about the retracted Brain study paperhttp://www.pnas.org/content/111/26/9657.abstractIdentify GO terms associated with a phenotypeMemory task was studied using fMRI and whole genome genotypingSNPs were mapped to genes (association P value for SNP calculated)Genes were ranked based on these P-values, gene-sets identifiedgene – go annotations were identifiedProblem was they flagged 11 genes using the same SNP (counted one variant 11 times)