Gowinda: unbiased analysis of gene set enrichment for genome-wide association studies

• Published in: Bioinformatics 2012-08, Vol.28 (15), p.2084-2085
• Main Authors: KOFLER, Robert, SCHLÖTTERER, Christian
• Format: Article
• Language: English
• Subjects: Animals; Applications Note; Biological and medical sciences; Computational Biology - methods; Drosophila melanogaster - genetics; Fundamental and applied biological sciences. Psychology; General aspects; Genome-Wide Association Study - methods; Linkage Disequilibrium; Mathematics in biology. Statistical analysis. Models. Metrology. Data processing in biology (general aspects); Polymorphism, Single Nucleotide; Probability; Software
• ISSN: 1367-4803; 1367-4811; 1460-2059
• DOI: 10.1093/bioinformatics/bts315

An analysis of gene set [e.g. Gene Ontology (GO)] enrichment assumes that all genes are sampled independently from each other with the same probability. These assumptions are violated in genome-wide association (GWA) studies since (i) longer genes typically have more single-nucleotide polymorphisms resulting in a higher probability of being sampled and (ii) overlapping genes are sampled in clusters. Herein, we introduce Gowinda, a software specifically designed to test for enrichment of gene sets in GWA studies. We show that GO tests on GWA data could result in a substantial number of false-positive GO terms. Permutation tests implemented in Gowinda eliminate these biases, but maintain sufficient power to detect enrichment of GO terms. Since sufficient resolution for large datasets requires millions of permutations, we use multi-threading to keep computation times reasonable. Gowinda is implemented in Java (v1.6) and freely available on http://code.google.com/p/gowinda/ [e-mail] Manual: http://code.google.com/p/gowinda/wiki/Manual. Test data and tutorial: http://code.google.com/p/gowinda/wiki/Tutorial. http://code.google.com/p/gowinda/wiki/VALIDATION.