getTermSim: Get pairwise GO term similarities.
In GOSim: Computation of functional similarities between GO terms and gene products; GO enrichment analysis

Description Usage Arguments Details Value Note Author(s) References See Also Examples

Returns the pairwise similarities between GO terms. Different calculation method are implemented.

1	getTermSim(termlist, method = "relevance", verbose = FALSE)

`termlist`	character vector of GO terms
`method`	one of the supported methods for GO term similarity (see below)
`verbose`	print out various information or not

Currently the following methods for computing GO term similarities are implemented:

"Resnik": information content of minimum subsumer (ICms) [1], here additionally divided by the maximum information content of all GO terms
"JiangConrath": 1 - \min(1, IC(term1) - 2ICms + IC(term2)) [2]
"Lin": \frac{2ICms}{(IC(term1) + IC(term2))} [3]
"CoutoEnriched": FuSSiMeg enriched term similarity by Couto et al. [4]. Requires enrichement factors to be set by setEnrichmentFactors

"CoutoResnik": average information content of common disjunctive ancestors of term1 and term2 (ICshare) [5]
"CoutoJiangConrath": 1 - \min(1, IC(term1) -2ICshare + IC(term2)) [5]
"CoutoLin": \frac{2ICshare}{(IC(term1) + IC(term2))} [5]
"diffKernel": diffusion kernel similarity from a pre-loaded kernel matrix (see load.diffusion.kernel). The diffusion kernel is calculated using one of the methods described in [6].
"relevance": sim_Lin * (1 - exp(-ICms)) [7]
"GIC": summed information content of common ancestors divided by summed information content of all ancestors of term1 and term2 [8]

n x n matrix (n = number of GO terms) with similarities between GO terms scaled to [0,1]. If a GO term does not exist for the currently set ontology, the similarity is set to "NA".

All calculations use normalized information contents for each GO term. Normalization is achieved by dividing each information content by the maximum information content within the currently set ontology ("BP","MF","CC")

Holger Froehlich

[1] P. Resnik, Using Information Content to evaluate semantic similarity in a taxonomy, Proc. 14th Int. Conf. Artificial Intel., 1995

[2] J. Jiang, D. Conrath, Semantic Similarity based on Corpus Statistics and Lexical Taxonomy, Proc. Int. Conf. Research in Comp. Ling., 1998

[3] D. Lin, An Information-Theoretic Definition of Similarity, Proc. 15th Int. Conf. Machine Learning, 1998

[4] F. Couto, M. Silva, P. Coutinho, Implementation of a Functional Semantic Similarity Measure between Gene-Products, DI/FCUL TR 03-29, Department of Informatics, University of Lisbon, 2003

[5] Couto, F.; Silva, M. & Coutinho, P., Semantic Similarity over the Gene Ontology: Family Correlation and Selecting Disjunctive Ancestors, Conference in Information and Knowledge Management, 2005

[6] Lerman G. & Shaknovich B., Defining Functional Distance using Manifold Embeddings of Gene Ontology Annotations, PNAS, 104(27): 11334 - 11339, 2007

[7] A. Schlicker, F. Domingues, J. Rahnenfuehrer, T. Lengauer, A new measure for functional similarity of gene products based on Gene Ontology, BMC Bioinformatics, 7, 302, 2006.

[8] C. Pesquita, D. Faria, H. Bastos, A. Falcao, F. Couto, Evaluating GO-based Semantic Similarity Measures, In: Proc. 10th Annual Bio-Ontologies Meeting 2007, 37 - 40, 2007

getMinimumSubsumer, getDisjCommAnc, setEnrichmentFactors, setOntology, load.diffusion.kernel

#\donttest{
 setOntology("BP")
 # Lin's method
 getTermSim(c("GO:0006955","GO:0007584"),method="Lin") 
# Couto's method combined with Jiang-Conrath distance
getTermSim(c("GO:0006955","GO:0007584"),method="CoutoJiangConrath") 

# set enrichment factors
setEnrichmentFactors(alpha=0.1,beta=0.5) 
getTermSim(c("GO:0006955","GO:0007584"),method="CoutoEnriched") 
#}