Uso da base de dados decundária KOG como ferramenta para caracterização de expressão gênica e mineração de dados em projetos transcriptoma

Abstract

A set of new results, techniques and tools are presented in this work for datamining and to help in the analysis of Expressed Sequence Tags (EST)generated by transcriptome projects. The secondary database KOG wasutilized as a tool in the alignment and automatic annotation of ESTs from fourorganisms, A. thaliana, C. elegans, D. melanogaster e H. sapiens. Thesealignments were utilized to many purposes and in many experiments, like:inference of similarity cutoffs utilizing tBLASTn with ESTs and proteins from the same organism; development of an annotation test with EST and KOGproteins; evaluation of the quality of annotation by using the cutoff valuesdiscovered; evaluation of the quality of annotation by using uniques generatedby the TGICL software; functional characterization of ESTs with KOG; evaluation of KOG coverage with incremental EST number and inference of aminimal number of EST to cover it; creation of a web tool named K-EST thatmakes available the EST sampling data with KOG and also the conservationdata among KOG clusters; inference of gene loss, or at least loss of geneexpression in organisms belonging or not to the KOG database, by using ESTsampling data and conservation.