Chibucos annot go_final

1. Marcus C. Chibucos, Ph.D.OntologyEvidenceAnnotationArabidopsis thaliana ATPaseHMA4 zinc binding domainGO:0006829 : zinc ion transport (BP)GO:0005886 : plasma membrane (CC)GO:0005515 : protein binding (MF)Gene Annotation And Ontology

2. Outline of this talk2Background: the language of biology

3. Gene Ontology: overview, terms & structure

4. Annotating with GO and Evidence

5. Using annotation to facilitate your researchAbout screenshots in this talk3AmiGO web-based ontology browserhttp://amigo.geneontology.orgOBO-Edit stand-alone editorhttp://oboedit.org

6. What is annotation? Who is involved?Term confusion (what’s in a name?)Scale: the sea of dataControlled vocabularies & ontologiesThe Gene Ontology ConsortiumBackground: the language of biology4

7. Annotation5annotate – to make or furnish critical or explanatory notes or comment. (Merriam-Webster dictionary)genome annotation – the process of taking the raw DNA sequence produced by the genome-sequencing projects and adding the layers of analysis and interpretation necessary to extract its biological significance and place it into the context of our understanding of biological processes. (Lincoln Stein, PMID 11433356)Gene Ontology annotation – the process of assigning GO terms to gene products… according to two general principles: first, annotations should be attributed to a source; second, each annotation should indicate the evidence on which it is based. (http://www.geneontology.org)

8. Diverse parties involved6End-users, including various researchersSmall-scale laboratory projectsWhole genome sequencing projectsAnnotatorsFrom reading papers to computational analysis Ontology developersCreate terms that reflect scientific knowledgeMake interoperable ontologies, database linksDevelopers of tools & resourcesStandards for storing & sharing dataWeb interfaces for data analysis & sharingMany areas of expertiseLaboratory sciences – biology, chemistry, medicine, and many other disciplinesComputational science – bioinformatics, genomics, statisticsSoftware development & web designPhilosophy – ontology & logic

9. Term confusion: synonyms7Do biologists use precise & consistent language?Mutually understood concepts – DNA, RNA, or proteinSynonym (one thing known by more than one name) – translation and protein synthesisEnzyme Commission reactionsStandardized id, official name & alternative nameshttp://www.expasy.ch/enzyme/2.7.1.40

10. Term confusion: homonyms8Homonyms common in biology – different things known by the same nameSporulationVascular (plant vasculature, i.e. xylem & phloem, or vascular smooth muscle, i.e. blood vessels?)Endospore formation Bacillus anthracis“Sporulation”Reproductive sporulationAsci & ascospores, Morchellaelata(morel)http://www.microbelibrary.org/ASMOnly/details.asp?id=1426&Lang=©L Stauffer 2003 (accessed 17-Sep-09)http://en.wikipedia.org/wiki/File:Morelasci.jpg©PG Warner 2008 (accessed 17-Sep-09)

11. Term confusion: homonyms and biological complexity 9AmiGO query “vascular”  51 termsIn biology, many related phenomena are described with similar terminology

12. The problem of scale10Small data sets, small experiments & isolated scientific communities?

13. Enormous data sets

14. Microarray experiments

15. Whole genome sequencing projects

16. Comparative genomics of multiple diverse taxa

17. Computers don’t understand nuance

18. Millions of proteins to annotate

19. How to effectively search?

20. How to draw meaningful comparisons?http://en.wikipedia.org/wiki/File:Microarray2.gif(accessed 17-Sep-09)

21. The Gene Ontology (GO)11Way to address the problems of synonyms, homonyms, biological complexity, increasing glut of dataGO provides a common biological language for protein functional annotationwww.geneontology.org

22. Controlled vocabulary (CV)12An official list of precisely defined terms that can be used to classify information and facilitate its retrievalThink of flat list like a thesaurus or catalog Benefits of CVsAllow standardized descriptions of thingsRemedy synonym & homonym issuesCan be cross-referenced externallyFacilitate electronic searchingA CV can be “…used to index and retrieve a body of literature in a bibliographic, factual, or other database. An example is the MeSH controlled vocabulary used in MEDLINE and other MEDLARS databases of the NLM.”http://www.nlm.nih.gov/nichsr/hta101/ta101014.html

23. Ontology is a type of CV with defined relationships13Ontology – formalizes knowledge of a subject with precise textual definitionsNetworked terms where child more specific (“granular”) than parentLess specificGO terms describe biological attributes of gene products…More granular

24. How GO works14GO Consortium develops & maintains:Ontologies and cross-links between ontologies and different resourcesTools to develop and use the ontologiesSourceForge tracker for developmentPeople studying organisms at databases annotate gene products with GO termsGroups share files of annotation data about their respective organismsBecause a common language was used to describe gene products and this information was shared amongst databases…We can search uniformly across databasesDo comparative genomics of diverse taxa

25. GO on SourceForgesourceforge.net/projects/geneontology15

26. The Gene Ontology Consortium16Collaboration began 1998 among model organism databases mouse (MGI), fruit fly (FlyBase) and baker’s yeast (SGD)Michael Ashburner of FlyBase contributed the base vocabularyToday > 20 members & associatesFirst publication 2000 (PMID 10802651)Today, PubMed query “gene ontology” yields 3,347 papers (27-Jun-2011)Organisms represented by GO annotations from every kingdom of lifeMany groups use GO in many different ways for their researchAmong eight OBO-Foundry ontologiesZFINReactomeIGS

27. OBO Foundry ontologieswww.obofoundry.org17Collaboration among developers of science-based ontologiesEstablish principles for ontology developmentGoal of creating a suite of orthogonal interoperable reference ontologies in the biomedical domain.many others…

28. What the GO is notGO comprises three ontologiesAnatomy & storage of GO termsOntology structureDetail of a term in AmiGOTrue path ruleGene Ontology:overview, terms & structure18

29. Caveats – what GO is not19Not gene naming system or gene catalogGO describes attributes of biological objects – “oxidoreductase activity” not “cytochromec”The three ontologies have limitationsNo sequence attributes or structural featuresNo characteristics unique to mutants or diseaseNo environment, evolution or expressionNo anatomy features above cellular componentNot dictated standard or federated solutionDatabases share annotations as they see fitCurators evaluate differentlyGO is evolving as our knowledge evolvesNew terms added on daily basisIncorrect/poorly defined terms made obsoleteSecondary ids – terms with same meaning merged

30. GO comprises three ontologies20Cellular component ontology (CC) “cytoplasm”Molecular function ontology (MF)“protein binding”“peptidase activity”“cysteine-type endopeptidase activity”Biological process ontology (BP)“proteolysis”“apoptosis”Terms describe attributes of gene products (GPs)Any protein or RNA encoded by a geneSpecies-independent context, e.g. “ribosome”Could describe GPs found in limited taxa, e.g. “photosynthesis” or “lactation”One GP can be associated with ≥ 1 CC, BP, MFExample: Caspase-6 from Bostaurus

31. Cellular component ontology21Describes location at level of subcellular structure & macromolecular complexGP subcomponent of or located in particular cellular component, with some exceptions:No individual proteins or nucleic acidsNo multicellular anatomical termsFor annotation purposes, a GP can be associated with or located in ≥ one cellular componentMulti-subunit enzyme or protein complex

32. ribosome

33. proteasome

34. ubiquitinligase complex

35. Anatomical structure

36. rough endoplasmic reticulum

37. nucleus

38. nuclear inner membraneMolecular function ontology22Describe gene product activity at molecular levelDescribes attributes of entitiesAdenylate cyclase (E.C. 4.6.1.1)Catalyzes a specific reaction:ATP = 3',5'-cyclic AMP + diphosphateDescribed by the Gene Ontology term:“adenylate cyclase activity” (GO:0004016)http://www.ebi.ac.uk/pdbsum/1ab8[accessed 4-Feb-2010]Usually single GP, sometimes a complex

39. “ferritin receptor activity”

40. Definition: “combining with ferritin, an iron-storing protein complex, to initiate a change in cell activity”

41. Broad functions

42. “catalytic activity”

43. “transporter activity”

44. “binding”

45. Specific functions

46. “adenylatecyclase activity”

47. “protein-DNA complex transmembrane transporter activity”

48. “Fc-gamma receptor I complex binding”Biological process ontology23Describes recognized series of events or molecular functions with a defined beginning and end“GO does not try to represent the dynamics or dependencies that would be required to fully describe a pathway” (from GO documentation)Mutant phenotypes often reflect disruptions in BPSpecific process

49. “pyrimidine metabolism”

50. “α-glucosidase transportGeneral considerationsThe Cell CycleThe Development NodeMulti-Organism ProcessMetabolismRegulationDetection of and Response to StimuliSensory PerceptionSignaling PathwaysTransport and LocalizationTransporter activity (molecular function)Other Misc. Standard DefsBroad process

51. “cellular physiological process”

52. “signal transduction”http://www.geneontology.org/GO.process.guidelines.shtml

53. Anatomy of a GO term24Term namegoid (unique numerical identifier)Synonyms (broad or narrow) for searching, alternative names, misspellings… Precise textual definition with reference stating sourceGO slimOntology placement

54. Storage and cross referencing of GO terms25Storage in flat file (text)

55. Database cross reference for mappings to GO

56. GO term identical to object in other databaseOntology structure:parent-child relationship26Parent term (broader)Child term (specialized)hexose metabolismmonosaccharide biosynthesishexose biosynthesisUp in the tree is more general; down in the tree is more specific:

57. Annotation of genes

58. Start with terms denoting broad functional categories

59. Use more specific term as knowledge warrantsOntology structure:terms arranged in DAGs27GO terms structured as hierarchical-like directed acyclic graphs (DAGs)Tree-like, but each term can have more than one parent (pseudo-hierarchy)Each term may have one or more child terms (“siblings” share same parent)parentschild termparentchild terms“siblings”

60. GO has three term relationships28is_a - child is instance of parent (“A is_a B”)Class-subclass relationshippart_of - child part of parent (“C part_of D”)When C present, part of D; but C not always presentNucleus always part_of cell; not all cells have nucleiregulatesChild term regulates parent term(Zoomed in view of biological process ontology depicted here.)

61. AmiGO for viewing terms29Open source HTML-based application developed by the GO ConsortiumInterface for browsing, querying and visualizing OBO dataUsers can search GO terms or annotationsAvailable via website or download for local installhttp://amigo.geneontology.orgExample query withkeyword “hemolysis” or goid GO:0019836GO:0019836

62. AmiGO search results30Click

63. Term information in AmiGO31Webpage continues…

64. AmiGO view continued32Several informative viewsClickNumber of gene products in GO annotation collection annotated to that term or one of its child termsRelationship between term and its parentOur term is much further down…

65. Graph view33Alternative view of network of termsA term with two parents34amine groupcarboxylic acid groupgeneric amino acidName: amino acid transmembrane transporter activity

66. ID number: GO:0015171

67. Definition: Catalysis of the transfer of amino acids from one side of a membrane to the other. Amino acids are organic molecules that contain an amino group and a carboxyl group. [source: GOC:ai, GOC:mtg_transport, ISBN:0815340729]

68. parent term: amine transmembrane transporter activity (GO:0005275)

69. relationship to parent: “is_a”

70. parent term: carboxylic acid transmembrane transporter activity (GO:0046943)

71. relationship to parent: “is_a”Multiple paths to root:graphical view in OBO-Edit35

72. “True path rule”36The pathway from a term all the way up to its top-level parent(s) must always be true for any gene product that could be annotated to that term (“if true for the child, then true for the parent”)Incorrect for Bacteriacell organellemitochondrion proton-transporting ATP synthase complexCorrect for Bacteria (and Eukaryotes)cell intracellular proton-transporting ATP synthase complex plasma membrane proton-transporting ATP synthase complex mitochondrial proton-transporting ATP synthase complex membrane plasma membrane plasma membrane proton-transporting ATP synthase complex organelle mitochondrion mitochondrial inner membrane mitochondrial proton-transporting ATP synthase complex(Abbreviated versions of the actualtrees)

73. What is GO annotation?Literature curation at model organism databasesThe annotation fileEvidence – critical for annotationSequence similarity-based annotationAnnotation specificityAnnotating with GO and Evidence37

74. GO annotation overview38Associating a GO term with a gene productGoal is to select GO terms from all three ontologies to represent what, where, and howLinking a GO term to a gene product asserts that it has that attributeFor example, 6-phosphofructokinaseMolecular functionGO:0003872 6-phosphofructokinase activityBiological processGO:0006096 glycolysisCellular componentGO:0005737 cytoplasmAnnotation, whether based on literature or computational methods, always involves:Learning something about a gene productSelecting an appropriate GO termProviding an appropriate evidence codeCiting a [preferably open access] referenceEntering information into GO annotation file

75. Chaperone DnaK, one protein/multiple annotations39Molecular functionATP binding (GO:0005524)ATPase activity (GO:0016887)unfolded protein binding (GO:0051082)misfolded protein binding (GO:0051787)denatured protein binding (GO:0031249)Biological processprotein folding (GO:0006457)protein refolding (GO:0042026)protein stabilization (GO:0050821)response to stress (GO:0006950)Cellular componentcytoplasm (GO:0005737)

76. Literature curation performed at model organism databases40From the abstract:

77. Results section indicates a “direct assay” annotation41They document the findings of a direct assay performed on purified protein:They further document the methods used, and evaluate the findings in the Discussion section…

78. Query AmiGO with “DNA ligase” & “DNA ligation”42All “ligation” in biological process ontology

79. Resulting annotations43Name: DNA ligase (stated in paper)Gene symbol: ligA (stated in paper)EC: 6.5.1.2 (queried enzyme for “DNA ligase”)

80. Gene annotation file captures annotations44Evidence

81. Evidence45Essential to base annotation on evidenceConclusions more robust and traceableWith evidence, a GO annotation is standard operating procedure (SOP)-independentMany types of evidence existFor example, experiment described in literatureWhat method (e.g. direct assay, mutant phenotype, et cetera) was used?Did author cite references?Did author provide details of analyses?Perhaps you used a sequence-based methodWhat were the methods of manual curation?Give accession numbers of similar sequencesProvide any references describing methodsControlled vocabularies help here, too!

82. GO standard references46GO_REF:0000011 A Hidden Markov Model (HMM) is a statistical representation of patterns found in a data set. When using HMMs with proteins, the HMM is a statistical model of the patterns of the amino acids found in a multiple alignment of a set of proteins called the "seed". Seed proteins are chosen based on sequence similarity to each other. Seed members can be chosen with different levels of relationship to each other... GO_REF:0000011 A Hidden Markov Model (HMM) is a statistical representation of patterns found in a data set. When using HMMs with proteins, the HMM is a statistical model of the patterns of the amino acids found in a multiple alignment of a set of proteins called the "seed". Seed proteins are chosen based on sequence similarity to each other. Seed members can be chosen with different levels of relationship to each other. They can be members of a superfamily (ex. ABC transporter, ATP-binding proteins), they can all share the same exact specific function (ex. biotin synthase) or they could share another type of relationship of intermediate specificity (ex. subfamily, domain). New proteins can be scored against the model generated from the seed according to how closely the patterns of amino acids in the new proteins match those in the seed. There are two scores assigned to the HMM which allow annotators to judge how well any new protein scores to the model. Proteins scoring above the "trusted cutoff" score can be assumed to be part of the group defined by the seed. Proteins scoring below the "noise cutoff" score can be assumed to NOT be a part of the group. Proteins scoring between the trusted and noise cutoffs may be part of the group but may not. One of the important features of HMMs is that they are built from a multiple alignment of protein sequences, not a pairwise alignment. This is significant, since shared similarity between many proteins is much more likely to indicate shared functional relationship than sequence similarity between just two proteins. The usefulness of an HMM is directly related to the amount of care that is taken in chosing the seed members, building a good multiple alignment of the seed members, assessing the level of specificity of the model, and choosing the cutoff scores correctly. In order to properly assess what functional relevance an above-trusted scoring HMM match has to a query, one must carefully determine what the functional scope of the HMM is. If the HMM models proteins that all share the same function then it is likely possible to assign a specific function to high-scoring match proteins based on the HMM. If the HMM models proteins that have a wide variety of functions, then it will not be possible to assign a specific function to the query based on the HMM match, however, depending on the nature of the HMM in question, it may be possible to assign a more general (family or subfamily level) function. In order to determine the functional scope of an HMM, one must carefully read the documentation associated with the HMM. The annotator must also consider whether the function attributed to the proteins in the HMM makes sense for the query based on what is known about the organism in which the query protein resides and in light of any other information that might be available about the query protein. After carefully considering all of these issues the annotator makes an annotation.

83. GO evidence codeswww.geneontology.org/GO.evidence.shtml47EXP - inferred from experimentIDA - inferred from direct assayIEP inferred from expression patternIGI - inferred from genetic interactionIPI - inferred from physical interactionIMP - inferred from mutant phenotypeISS - inferred from sequence or structural similarityISA - inferred from sequence alignmentISO - inferred from sequence orthologyISM - inferred from sequence modelIGC - inferred from genomic contextND - no biological data availableIC - inferred by curatorTAS - traceable author statementNAS - non-traceable author statementIEA - inferred from electronic annotationGO codes are a subset of yet another ontology!

84. Types of sequence similarity-based annotations48Find similarity between gene product & one that is experimentally characterizedBLAST-type alignmentsShared synteny to establish orthology of genomic regions between speciesFind similarity between gene product and defined protein familyHMMs (Pfam, TIGRFAMS)PrositeInterProFind motifs in gene product with prediction toolsTMHMM SignalPMany (most?) information you find is based on transitive annotation and much of it has never been looked at by a human being!

85. Evaluation of sequence similarity-based information49Visually inspect alignments & criteriaLength & identityConservation of catalytic sitesCheck HMM scores with respect to cutoffLook at available metabolic analysisPathways, complexes?Information from neighboring genesGene in an operon (common prokaryotes) can supplement weak similarity evidenceSequence characteristicsTransmembraneregions?Signal peptide?Known motifs that give a clue to function?Paralogous family member

86. An example: HI0678, a protein from H. influenzae…...high quality alignment to experimentally characterized triosephosphateisomerase from Vibrio marinus50

87. Information from Swiss-Prot database on experimentally characterized match proteinfurther down the page51

88. High quality…..…. full-length match, high percent identity (67.8%), conserved active and binding sites (boxed in red).52

89. Resulting annotations53name:triosephosphateisomerasegene symbol:tpiAEC: 5.3.1.1(This, and the following annotations, came from the match protein.)

90. KEGG pathway for glycolysis core54

91. KEGG pathway for glycolysis core55

92. Resulting annotations56name: triosephosphateisomerasegene symbol: tpiAEC: 5.3.1.1

93. And another annotation57The biologist knows that glycolysis takes place in the cytoplasm in bacteria, and so infers a cytoplasmic location for that protein (“inferred by curator” evidence code).

94. Annotation specificity should reflect knowledge58GO trees (very abbreviated)Functioncatalytic activitykinase activity carbohydrate kinase activityribokinase activityglucokinase activityfructokinase activityProcess metabolism carbohydrate metabolism monosaccharide metabolismhexose metabolismglucose metabolism fructose metabolism pentose metabolism ribose metabolismAvailable evidence for three genes#1-good match to an HMM for “kinase”#2-good match to an HMM for “kinase”-a high-quality BER match to an experimentally characterized “glucokinase’ AND a ‘fructokinase’#3-good match to an HMM specific for “ribokinase”-a high-quality BER match to an experimentally characterized ribokinase#1#2#3#1#2#3

95. Using shared annotationsSearch for GO terms at databasesSlims for broad classificationGO toolsWorking with GO-limited data setsSummaryUsing annotation to facilitate your research59

96. Sharing annotations60Annotation file sent to GO, put in repositoryAll these data free to anyoneHundreds of thousands of GP annotationsAnnotation files all in same formatFacilitates easy use of data by everyoneMost of your favorite organism databases use these annotation files

97. Searching for GO terms at EuPathDB61

98. 62Ontology slimwww.geneontology.org/GO.slims.shtmlSlim is a distilled (reduced) ontology Made by manually pruning low-level terms with an ontology editorSelected high-level terms remainSlims reduce ontology complexityReduce clutter & see general trendsMicroarray experimentsComparative whole genome analysesRemove irrelevant termsLooking at specific taxa, such as yeast or plantGo offers script to bin more granular annotations up to higher levels

99. Comparing genomes with a GO slim63High-levelbiological process terms used to compare Plasmodium and SaccharomycesMJ Gardner, et al. (2002) Nature 419:498-511

100. GO slim: manual/orthology-based gene annotations64Nucleic Acids Res. 2010 January; 38(Database issue): D420–D427.

101. GO toolswww.geneontology.org/GO.tools.shtml65The real challenge is finding the right one for your needsFor example, statistical representation of GO terms:http://go.princeton.edu/cgi-bin/GOTermFinder

102. GO & analysis of RNA-seqdata66Young et al. Genome Biology 2010, 11:R14 http://genomebiology.com/2010/11/2/R14We present GOseq, an application for performing Gene Ontology (GO) analysis on RNA-seq data. GO analysis is widely used to reduce complexity and highlight biological processes in genome-wide expression studies, but standard methods give biased results on RNA-seq data due to over-detection of differential expression for long and highly expressed transcripts. Application of GOseq to a prostate cancer data set shows that GOseq dramatically changes the results, highlighting categories more consistent with the known biology.

103. When GO is limited67Food for thought: what happens when we have limited GO (or other)annotation data?New and interesting genomes often see this problem

104. Comparative analysis of orthologs in syntenic blocks68The more genomes we have at our disposal, the betterStructural rearrangements, absence of intron, gene duplication, intron structure, gene deletion/creationNucleic Acids Res. 2010 January; 38(Database issue): D420–D427.

105. Summary GO analyses69GO remedies problems of synonyms & homonyms in biological nomenclatureQueries based on IDs linked to precise definitions, not less reliable text-matchingGO can help you to:Find all genes that share a particular function regardless of sequenceDo comparisons across any species annotated with GOSummarize major classes of genes in a newly sequenced genomeCharacterize expressed genes is a studyDrive hypotheses to test in the laboratoryGO is not a panacea but it should be a valuable tool in your genomics toolbox

106. The title slide revisited…OntologyEvidenceAnnotationArabidopsis thaliana ATPaseHMA4 zinc binding domainGO:0006829 : zinc ion transport (BP)GO:0005886 : plasma membrane (CC)GO:0005515 : protein binding (MF)Thank you.

Chibucos annot go_final

More Related Content

What's hot (20)

Viewers also liked (9)

Similar to Chibucos annot go_final (20)

More from Sucheta Tripathy (20)

Recently uploaded (20)

Chibucos annot go_final

Editor's Notes