Emily Dimmer GOA group European Bioinformatics Institute Wellcome Trust Genome Campus Cambridge UK Gene Ontology (GO)

Introduction to GO Description of the GO ontologies How groups annotate to GO Practical: Investigating the GO and OBO web sites Browsing the GO using the AmiGO Browser. Open Biomedical Ontologies How GO is being used Available Tools GO slims Practical: Creating your own GO slim GO Tutorial Outline:

Introduction to GO Description of the GO ontologies How groups annotate to GO Practical: Investigating the GO and OBO web sites Browsing the GO using the AmiGO Browser. Open Biomedical Ontologies How GO is being used Available Tools GO slims Practical: Creating your own GO slim GO Tutorial Outline:

Introduction to GO Description of the GO ontologies How groups annotate to GO Practical: Investigating the GO and OBO web sites Browsing the GO using the AmiGO Browser. Open Biomedical Ontologies How GO is being used Available Tools GO slims Practical: Creating your own GO slim GO Tutorial Outline:

Introduction to GO Description of the GO ontologies How groups annotate to GO Practical: Investigating the GO and OBO web sites Browsing the GO using the AmiGO Browser. Open Biomedical Ontologies How GO is being used Available Tools GO slims Practical: Creating your own GO slim GO Tutorial Outline:

Why is GO needed ? THE PROBLEM: Huge body of knowledge with an extremely large vocabulary to describe it Vocabulary used is poorly defined –i.e. one word can have different meanings –or different names for the same concept Biological systems are complex and our knowledge of such systems is incomplete RESULT: Large databases which are difficult to manage and impossible to mine computationally

A (part of the) solution: GO: “a controlled vocabulary that can be applied to all organisms even as knowledge of gene and protein roles in cells is accumulating and changing” What is GO?

Access gene product functional information Provide a link between biological knowledge and … gene expression profiles proteomics data Find how much of a proteome is involved in a process/ function/ component in the cell using a GO-Slim (a slimmed down version of GO to summarize biological attributes of a proteome) Map GO terms and incorporate manual GOA annotation into own databases to enhance your dataset or to validate automated ways of deriving information about gene function (text-mining). What can scientists do with GO?

Tactition Tactile sense Taction ?

perception of touch ; GO: Tactition Tactile sense Taction

Molecular Function: elemental activity or task e.g. DNA binding, catalysis of a reaction Biological Process: broad objective or goal e.g. mitosis, signal transduction, metabolism Cellular Component: location or complex e.g. nucleus, ribosome GO Three (Orthogonal) Ontologies

Molecular Function: elemental activity or task e.g. DNA binding, catalysis of a reaction Biological Process: broad objective or goal e.g. mitosis, signal transduction, metabolism Cellular Component: location or complex e.g. nucleus, ribosome GO Three (Orthogonal) Ontologies

Molecular Function: elemental activity or task e.g. DNA binding, catalysis of a reaction Biological Process: broad objective or goal e.g. mitosis, signal transduction, metabolism Cellular Component: location or complex e.g. nucleus, ribosome GO Three (Orthogonal) Ontologies

Molecular Function: elemental activity or task e.g. DNA binding, catalysis of a reaction Biological Process: broad objective or goal e.g. mitosis, signal transduction, metabolism Cellular Component: location or complex e.g. nucleus, ribosome GO Three (Orthogonal) Ontologies

How does GO work? Provides a standard, species-neutral way of representing biology GO covers ‘normal’ functions and processes –No pathological processes –No experimental conditions

Molecular Function 7,493 terms Biological Process 9,640 terms Cellular Component 1,634 terms Total 18,767 terms Definitions: 16,696 (93.9 %) Content of GO

What is GO? NOT a system of nomenclature or a list of gene products GO doesn’t attempt to cover all aspects of biology or evolutionary relationships Open Biomedical Ontologies NOT a dictated standard NOT a way to unify databases

Reactome

Anatomy of a GO term GO terms are composed of: Term name Unique GO ID Definition (93 % of GO terms are defined) Synonyms (optional) Database references (optional) Relationships to other GO terms

Ontologies “Ontologies provide controlled, consistent vocabularies to describe concepts and relationships, thereby enabling knowledge sharing” (Gruber 1993) I. The GO Ontologies

Can be used to: Formalise the representation of biological knowledge Describe a common and defined vocabulary for database annotation Standardise database submissions Provide unified access to information through ontology-based querying of databases, both human and computational Improve management and integration of data within databases. Facilitate data mining Ontology applications

Ontologies can be represented as graphs, where the vertices (nodes and leaves) are connected by edges. The nodes are concepts in the ontology. The edges are the relationships between the concepts node edge Ontology Structure

The Gene Ontology is structured as a hierarchical directed acyclic graph (DAG). Terms are linked by two relationships –is-a –part-of Terms can have more than one parent

Simple hierarchies Directed Acyclic (Trees) Graphs

Directed Acyclic Graph cell membrane chloroplast mitochondrial chloroplast membrane is-a part-of

True Path Rule The path from a child term all the way up to its top-level parent(s) must always be true cell  cytoplasm  chromosome  nuclear chromosome  nucleus  nuclear chromosome is-a  part-of 

Terms become obsolete when they are removed or redefined GO IDs are never deleted For each term, a comment is added to explains why the term is now obsolete Ensuring Stability in a Dynamic Ontology Obsolete Cellular Component Obsolete Molecular Function Obsolete Biological Process Biological Process Molecular Function Cellular Component

Access to the Gene Ontology Downloads formats available: OBO GO XMLOWL MySQL ( Web-based tools AmiGO ( QuickGO (

II. Annotating to GO Use of GO terms to represent the activities and localizations of gene products. Basic information needed: 1. Database object (e.g. a protein or gene identifier) e.g. Q9ARH1 2. Reference ID e.g. PubMed ID: GO term ID e.g. GO: Evidence code e.g. TAS

GenNav:

J. Clark et al. Plant Physiology 2005 (in press)

Two types of GO Annotation:  Electronic Annotation  Manual Annotation All annotations must: be attributed to a source. indicate what evidence was found to support the GO term-gene/protein association.

Electronic Annotation Provides large-coverage High-quality BUT annotations tend to use high-level GO terms and provide little detail.

1.Assignment of GO terms to gene products using existing information within database entries Manual mapping of GO terms to concepts external to GO (‘translation tables’). Proteins then electronically annotated with the relevant GO term(s). 2.Automatic sequence analyses to transfer annotations between highly similar gene products Electronic Annotation

Fatty acid biosynthesis ( Swiss-Prot Keyword) EC: (EC number) IPR000438: Acetyl-CoA carboxylase carboxyl transferase beta subunit ( InterPro entry) MF_00527: Putative 3- methyladenine DNA glycosylase (HAMAP) GO:Fatty acid biosynthesis ( GO: ) GO:acetyl-CoA carboxylase activity ( GO: ) GO:acetyl-CoA carboxylase activity (GO: ) GO:DNA repair (GO: ) Electronic Annotation

Mappings of external concepts to GO

Evaluation of precision of annotation electronic techniques (InterPro2GO, SPKW2GO, EC2GO) Compared manually-curated test set of GO annotated proteins with the electronic annotations InterPro2GO = most coverage EC2GO = 67 % of predictions exactly match the manual GO annotation % of time the 3 mappings predicted GO terms within the same lineage Camon et al. BMC Bioinformatics 2005 in press

Manual Annotation High–quality, specific gene/gene product associations made, using: Peer-reviewed papers Evidence codes to grade evidence BUT – is very time consuming and requires trained biologists

Finding GO terms In this study, we report the isolation and molecular characterization of the B. napus PERK1 cDNA, that is predicted to encode a novel receptor-like kinase. We have shown that like other plant RLKs, the kinase domain of PERK1 has serine/threonine kinase activity, In addition, the location of a PERK1-GTP fusion protein to the plasma membrane supports the prediction that PERK1 is an integral membrane protein…these kinases have been implicated in early stages of wound response… Process: response to wounding GO: serine/threonine kinase activity, Function: protein serine/threonine kinase activity GO: integral membrane protein Component: integral to plasma membrane GO: …for B. napus PERK1 protein (Q9ARH1) PubMed ID: wound response

GO Evidence Codes *With column required Manually annotated CodeDefinition *IEAInferred from Electronic Annotation IDAInferred from Direct Assay IEPInferred from Expression Pattern *IGIInferred from Genetic Interaction IMPInferred from Mutant Phenotype *IPIInferred from Physical Interaction *ISSInferred from Sequence Similarity TASTraceable Author Statement NASNon-traceable Author Statement *ICInferred from Curator RCAInferred from Reviewed Computational Analysis NDNo Data IDA: Enzyme assays In vitro reconstitution (transcription) Immunofluorescence Cell fractionation TAS: In the literature source the original experiments referred to are traceable (referenced).

GO Evidence Codes *With column required Manually annotated additional needed identifier for annotations using certain evidence codes CodeDefinition *IEAInferred from Electronic Annotation IDAInferred from Direct Assay IEPInferred from Expression Pattern *IGIInferred from Genetic Interaction IMPInferred from Mutant Phenotype *IPIInferred from Physical Interaction *ISSInferred from Sequence Similarity TASTraceable Author Statement NASNon-traceable Author Statement *ICInferred from Curator RCAInferred from Reviewed Computational Analysis NDNo Data IGI: a gene identifier for the "other" gene involved in the interaction IPI: a gene or protein identifier for the "other" protein involved in the interaction IC: GO term from another annotation used as the basis of a curator inference

Annotation of a gene product to one ontology is independent from its annotation to other ontologies. Terms reflecting a normal activity or location are only annotated to. Usage of ‘unknown’ GO terms (e.g. Molecular function unknown GO: ) …some extra things:

A set of ‘Qualifier’ terms is also available to curators modify the interpretation of an annotation. Allowable values: 1. NOT a gene product is not associated with the GO term to document conflicting claims in the literature. 2. Contributes to distinguishes between individual subunits functions and whole complex functions (used with GO Function Ontology) 3. Colocalizes with Transiently or peripherally associated with an organelle or complex where the resolution of an assay is not accurate. (used with GO Component Ontology) …some extra things: Qualifier Information

The Qualifier column can be used to modify the interpretation of an annotation. Allowable values: 1. NOT a gene product is not associated with the GO term to document conflicting claims in the literature. 2. Contributes to distinguishes between individual subunits functions and whole complex functions (used with GO Function Ontology) 3. Colocalizes with Transiently or peripherally associated with an organelle or complex where the resolution of an assay is not accurate. (used with GO Component Ontology) …some extra things:

The Qualifier column can be used to modify the interpretation of an annotation. Allowable values: 1. NOT a gene product is not associated with the GO term to document conflicting claims in the literature. 2. Contributes to distinguishes between individual subunits functions and whole complex functions (used with GO Function Ontology) 3. Colocalizes with Transiently or peripherally associated with an organelle or complex where the resolution of an assay is not accurate. (used with GO Component Ontology) …some extra things:

The Qualifier column can be used to modify the interpretation of an annotation. Allowable values: 1. NOT a gene product is not associated with the GO term to document conflicting claims in the literature. 2. Contributes to distinguishes between individual subunit functions and whole complex functions (used with GO Function Ontology) 3. Colocalizes with Transiently or peripherally associated with an organelle or complex where the resolution of an assay is not accurate. (used with GO Component Ontology) …some extra things:

Accessing annotations to the Gene Ontology 1. Downloads Annotations – gene association files Ontologies and annotations – MySQL and XML 2. Web-based access AmiGO ( QuickGO ( …among others…

Gene Association File Calcyclin IPI protein taxon: UniProt Calcyclin IPI protein taxon: UniProt UniProtP06703S106_HUMAN GO: GOA:spkw IEA F UniProtP06703 S106_HUMAN NOT GO: PMID: NAS P UniProtP06703 S106_HUMAN GO: PMID: IPI UniProt:P50995 F via web (GO consortium page) DB DB_Object_ID DB_Object_Symbol Qualifier GOid DB:Reference Evidence With Aspect DB_Object_Name DB_Object_Synonym DB_Object_Type taxon Date Assigned by

Summary GO is still being developed and updated - it requires a serious and ongoing effort. –the biological community is involved New model organism databases are joining the GO Consortium annotation effort

Practical session 1.Visit the GO website 2.Visit the OBO website 3.Browse the ontologies using the official GO Consortium Browser – AmiGO

GO web site: Part 1.

OBO web site:

AmiGO:

GO terms with no children

Filter queries by organism, data source or evidence Search for GO terms or by Gene symbol/name Querying the GO

GOst tool

QuickGO browser:

OBO and Gene Ontology Uses and Tools

Anatomy Physiology Phenotype Pathway Disease Molecular Metabolic Developmental Stage Ontologies

Beyond GO – Open Biomedical Ontologies Orthogonal to existing ontologies to facilitate combinatorial approaches - Share unique identifier space - Include definitions Anatomies Cell Types Sequence Attributes Temporal Attributes Phenotypes Diseases More….

Sequence Ontology

Ontology of ‘small molecular entities’

Access to GO and its annotations

How to access the Gene ontology and its annotations 1. Downloads Ontologies – (various – GO, OBO, XML, OWL MySQL) Annotations – gene association files Ontologies and Annotations – MySQL and XML 2. Web-based access AmiGO ( QuickGO ( among others…

SRS view…

Access gene product functional information Provide a link between biological knowledge and … gene expression profiles proteomics data Find how much of a proteome is involved in a process/ function/ component in the cell using a GO-Slim (a slimmed down version of GO to summarize biological attributes of a proteome) Map GO terms and incorporate manual GOA annotation into own databases to enhance your dataset or to validate automated ways of deriving information about gene function (text-mining). What can scientists do with GO?

attacked time control Puparial adhesion Molting cycle hemocyanin Defense response Immune response Response to stimulus Toll regulated genes JAK-STAT regulated genes Immune response Toll regulated genes Amino acid catabolism Lipid metobolism Peptidase activity Protein catabloism Immune response Bregje Wertheim at the Centre for Evolutionary Genomics, Department of Biology, UCL and Eugene Schuster Group, EBI. …analysis of high-throughput data according to GO MicroArray data analysis

Proteomics data analysis Kislinger T et al, Mol Cell Proteomics, 2003 GO classification …analysis of high-throughput data according to GO

Analysis of Data: Clustering

Color indicates up/down regulation GoMiner Tool, John Weinstein et al, Genome Biol. 4 (R28) 2003

Compare annotations associated with the test set to the entire set of GO annotations…. DNA Repair seems to be a common theme. Example of VLAD Output

…overview proteome with GO Slim

map2slim.pl distributed as part of the go-perl package maps a set of annotations up to their parent GO slim terms Off-the-shelf GO slims

Summary  The Gene Ontology project precipitated a generalized implementation for ontologies for molecular biology  Bio-ontologies such as GO have facilitated development of systems for hypothesis generation in biological systems  Further integration – creation of cross-products between different ontologies

Practical II – Creation of GO slims using the DAG-Edit tool.

…loading the GO

ftp://ftp.geneontology.org/pub/go/ontology/gene_ontology.obo …loading the GO

…browsing the GO

…viewing GO terms

…searching for GO terms

…creating a new GO slim

…creating a renderer for the GO slim

…adding terms to the GO slim

…filtering GO for terms in the GO slim

…removing filters/renderers

…saving the newly created GO slim