||November 16, 2009
||Dr. Avraham A. Melkman, Computer Science Department, Ben Gurion University, Israel
|| Integrating gene annotations into data analysis
Considerable effort has been invested in collecting information about genes in the form of annotation. Our research revolves around using parts of this information to improve various methods of data analysis. We have concentrated in particular on the gene annotations provided by the Gene Ontology.
The talk will describe two projects that draw especially upon the hierarchical structure of the Gene Ontology. The purpose of the first is to concisely characterize a long list of annotation terms, a typical product of an enrichment analysis of a study-set of genes.
The second utilizes the hierarchical structure through its contribution to the (semantic) similarity between annotation terms. The purpose here is to improve a clustering of the genes, based for example on expression data, by adjusting the clustering to take into account the additional knowledge embodied in the annotations of the genes and their semantic similarities.
Both works are joint with Dikla Dotan-Cohen, and the second is also joint with Simon Kasif.