1. Introduction to the Gene Ontology
Genomic Annotation and Functional Modeling Workshop
Maxwell H. Gluck Equine Research Center
15-16 November, 2011

2. Introduction to GO The Gene Ontology Consortium The Gene ontology
A GO annotation example
GO evidence codes
no GO vs ND
Making Annotations
Multiple annotations - the gene association (ga) file
Sources of GO

3. The Gene Ontology Consortium

5. The GO Consortium provides:
central repository for ontology updates and annotations
central mechanism for changing GO terms (adding, editing, deleting)
quality checking for annotations
consistency checks for how annotations are made by different groups
central source of information for users
co-ordination of annotation effort

6. GO Consortium and GO Groups:
groups decide gene product set to annotate
biocurator training
tool development mostly by groups
many non-consortium groups
education and training by groups
outreach to biocurators/databases by GOC

7. Annotation Strategy Experimental data Computational analysis
Many species have a body of published, experimental data
Detailed, species-specific annotation: ‘depth’
Requires manual annotation of literature  slow
Computational analysis
Can be automated  faster
Gives ‘breadth’ of coverage across the genome
Annotations are general
Relatively few annotation pipelines

8. Releasing GO Annotations
GO annotations are stored at individual databases
Sanity checks as data is entered – is all the data required filled in?
Databases do quality control (QC) checks and submit to GO
GO Consortium runs additional QC and collates annotations
Checked annotations are picked up by GO users
eg. public databases, genome browsers, array vendors, GO expression analysis tools

9. AgBase Quality Checks & Releases
AgBase Biocurators
‘sanity’ check
AgBase
biocuration
interface
‘sanity’ check
& GOC QC
AgBase database
GO analysis tools
Microarray developers
‘sanity’ check
UniProt db
QuickGO browser
GO analysis tools
Microarray developers
EBI GOA Project
‘sanity’ check: checks to ensure all appropriate information is captured, no obsolete GO:IDs are used, etc.
‘sanity’ check
& GOC QC
Public databases
AmiGO browser
GO analysis tools
Microarray developers
GO Consortium
database

10. The Gene Ontology

11. Gene Ontology (GO) Not about genes! Not a single ontology
Gene products: genes, transcripts, ncRNA, proteins
The GO describes gene product function
Not a single ontology
Biological Process (BP or P)
Molecular Function (MF or F)
Cellular Component (CC or C)
de facto method for functional annotation
Widely used for functional genomics (high throughput).

12. What the GO doesn’t do: Does not describe individual gene products
e.g. cytochrome c is not in the GO but oxidoreductase activity is
Does not describe mutants or diseases, e.g. oncogenesis.
Does not include sequence attributes, e.g., exons, introns, protein domains.
Is not a database of sequences.

13. What is the Gene Ontology?
“a controlled vocabulary that can be applied to all organisms even as knowledge of gene and protein roles in cells is accumulating and changing”
assign functions to gene products at different levels, depending on how much is known about a gene product
is used for a diverse range of species
structured to be queried at different levels, eg:
find all the chicken gene products in the genome that are involved in signal transduction
zoom in on all the receptor tyrosine kinases
human readable GO function has a digital tag to allow computational analysis of large datasets

14. relationships between terms
Ontologies
relationships between terms
digital identifier
(computers)
description
(humans)
As of ontology version (30/09/2011)
35,029 terms, 100.0% defined
* 21,439 biological process
* 2,898 cellular component
* 9,107 molecular function
1,585 obsolete terms (not included in figures above)

15. A GO Annotation example

16. A GO Annotation Example
NDUFAB1 (UniProt P52505)
Bovine NADH dehydrogenase (ubiquinone) 1, alpha/beta subcomplex, 1, 8kDa
Biological Process (BP or P)
GO: fatty acid biosynthetic process TAS
GO: mitochondrial electron transport, NADH to ubiquinone TAS
GO: lipid biosynthetic process IEA
Molecular Function (MF or F)
GO: fatty acid binding IDA
GO: NADH dehydrogenase (ubiquinone) activity TAS
GO: oxidoreductase activity TAS
GO: acyl carrier activity IEA
NDUFAB1
Cellular Component (CC or C)
GO: mitochondrial matrix IDA
GO: mitochondrial respiratory chain complex I IDA
GO: mitochondrion IEA

17. A GO Annotation Example
NDUFAB1 (UniProt P52505)
Bovine NADH dehydrogenase (ubiquinone) 1, alpha/beta subcomplex, 1, 8kDa
aspect or ontology
GO:ID (unique)
GO term name
GO evidence code

18. GO Evidence codes & Making annotations

19. Why record GO evidence code?
GO did not initially record evidence for functional assertion:
NR: Not Recorded
“inferred from…”
deduce or conclude (information) from evidence and reasoning
provides information about the support for associating a gene product with a function
different experiments allow us to draw different conclusions
reliability

22. Types of GO Evidence Codes
Experimental Evidence Codes
Computational Analysis Evidence Codes
Author Statement Evidence Codes
Curator Statement Evidence Codes
Automatically-assigned Evidence Codes
Obsolete Evidence Codes

23. Guide to GO Evidence Codes
GO EVIDENCE CODES
Direct Evidence Codes
IDA - inferred from direct assay
IEP - inferred from expression pattern
IGI - inferred from genetic interaction
IMP - inferred from mutant phenotype
IPI - inferred from physical interaction
Indirect Evidence Codes
inferred from literature
IGC - inferred from genomic context
TAS - traceable author statement
NAS - non-traceable author statement
IC - inferred by curator
inferred by sequence analysis
RCA - inferred from reviewed computational analysis
IS* - inferred from sequence*
IEA - inferred from electronic annotation
Other
NR - not recorded (historical)
ND - no biological data available
ISS - inferred from sequence or structural similarity
ISA - inferred from sequence alignment
ISO - inferred from sequence orthology
ISM - inferred from sequence model

24. GO Mapping Example GO EVIDENCE CODES Direct Evidence Codes
IDA - inferred from direct assay
IEP - inferred from expression pattern
IGI - inferred from genetic interaction
IMP - inferred from mutant phenotype
IPI - inferred from physical interaction
Indirect Evidence Codes
inferred from literature
IGC - inferred from genomic context
TAS - traceable author statement
NAS - non-traceable author statement
IC - inferred by curator
inferred by sequence analysis
RCA - inferred from reviewed computational analysis
IS* - inferred from sequence*
IEA - inferred from electronic annotation
Other
NR - not recorded (historical)
ND - no biological data available
GO Mapping Example
Biocuration of literature
detailed function
“depth”
slower (manual)
NDUFAB1

25. Biocuration of Literature:
P05147
Biocuration of Literature:
detailed gene function
Find a paper
about the protein.
PMID:

26. Read paper to get experimental evidence of function
Use most specific term
possible
Read paper to get experimental evidence of function
experiment assayed kinase activity:
use IDA evidence code
Same piece of data (IDA) demonstrates that this gene product inhibit protein kinase activity and thus is involved in the negative regulation of protein amino acid phosphorylation

27. GO Mapping Example GO EVIDENCE CODES Direct Evidence Codes
IDA - inferred from direct assay
IEP - inferred from expression pattern
IGI - inferred from genetic interaction
IMP - inferred from mutant phenotype
IPI - inferred from physical interaction
Indirect Evidence Codes
inferred from literature
IGC - inferred from genomic context
TAS - traceable author statement
NAS - non-traceable author statement
IC - inferred by curator
inferred by sequence analysis
RCA - inferred from reviewed computational analysis
IS* - inferred from sequence*
IEA - inferred from electronic annotation
Other
NR - not recorded (historical)
ND - no biological data available
GO Mapping Example
Biocuration of literature
detailed function
“depth”
slower (manual)
Sequence analysis
rapid (computational)
“breadth” of coverage
less detailed
NDUFAB1
ISS - inferred from sequence or structural similarity
ISA - inferred from sequence alignment
ISO - inferred from sequence orthology
ISM - inferred from sequence model

28. Computational Analysis Evidence
In the beginning:
IGC: Inferred from Genomic Context
e.g. operons
RCA: inferred from Reviewed Computational Analysis
computational analyses that integrate datasets of several types
ISS: Inferred from Sequence or Structural Similarity

29. Computational Analysis Evidence
Then different types of sequence analysis added:
ISS: Inferred from Sequence or Structural Similarity
ISO: Inferred from Sequence Orthology
ISA: Inferred from Sequence Alignment
ISM: Inferred from Sequence Model

30. Computational Analysis Evidence
Phylogenetic analysis codes added:
IBA: Inferred from Biological aspect of Ancestor
IBD: Inferred from Biological aspect of Descendant
IKR: Inferred from Key Residues
characterized by the loss of key sequence residues - implies a NOT annotation
IRD: Inferred from Rapid Divergence
characterized by rapid divergence from ancestral sequence – implies a NOT annotation
Exact details/use under discussion.

31. Unknown Function vs No GO
ND – no data
Biocurators have tried to add GO but there is no functional data available
Previously: “process_unknown”, “function_unknown”, “component_unknown”
Now: “biological process”, “molecular function”, “cellular component”
No annotations (including no “ND”): biocurators have not annotated
this is important for your dataset: what % has GO?

33. Multiple Annotations: gene association files

34. The gene association (ga) file
standard file format used to capture GO annotation data
tab-delimited file containing 15* fields of information:
Information about the gene product (database, accession, name, symbol, synonyms, species)
information about the function:
GO ID, ontology, reference, evidence, qualifiers, context (with/from)
data about the functional annotation
date, annotator
* GO Annotation File Format 2.0 has two additional columns compared to GAF 1.0: annotation extension (column 16) and gene product form ID (column 17).

36. (additional column added to this example)

37. gene product information

38. metadata: when & who

39. function information

40. Used to give more specific information about the evidence code
(not always displayed)

41. Used to qualify the annotation
(not always displayed)

43. Gene association files
GO Consortium ga files
many organism specific files
also includes EBI GOA files
EBI GOA ga files
UniProt file contains GO annotation for all species represented in UniProtKB
AgBase ga files
organism specific files
AgBase GOC file – submitted to GO Consortium & EBI GOA
AgBase Community file – GO annotations not yet submitted or not supported / annotations provided by researchers
all files are quality checked

49. Sources of GO
Primary sources of GO: from the GO Consortium (GOC) & GOC members
most up to date
most comprehensive
Secondary sources: other resources that use GO provided by GOC members
public databases (eg. NCBI, UniProtKB)
genome browsers (eg. Ensembl)
array vendors (eg. Affymetrix)
GO expression analysis tools

50. Sources of GO annotation
Different tools and databases display the GO annotations differently.
Since GO terms are continually changing and GO annotations are continually added, need to know when GO annotations were last updated.

51. Secondary Sources of GO annotation
EXAMPLES:
public databases (eg. NCBI, UniProtKB)
genome browsers (eg. Ensembl)
array vendors (eg. Affymetrix)
CONSIDERATIONS:
What is the original source?
When was it last updated?
Are evidence codes displayed?

52. Differences in displaying GO annotations: secondary/tertiary sources.

53. For more information about GO
GO Evidence Codes:
gene association file information:
tools that use the GO:
GO Consortium wiki:
All websites are listed on the AgBase workshop website.