ASSIGNING FUNCTION TO ESTs

John Innes Centre, Colney, Norwich NR4 7UH, UK

ESTs are the product of high through-put single pass sequencing of cDNAs, each being on average 400 bp. Because contaminating sequences and contaminating ESTs are not rigorously excluded from the EST libraries in dbEST care must be taken when interpreting the results of any analysis. ESTs provide a rapid way to obtain information about gene expression in specific cell types and they provide a means for gene discovery. Nearly 55% of all ESTs in dbEST derive from genes which have yet to be discovered and the function of the majority of these can not be predicted. Here we describe a method of automatic analysis using blastx which allows the association of ESTs with genes, gene families and protein domains in the absence of any quality control. Subsequent next-nearest-neighbor analysis can be used to identify singletons, orthologous and paralogous groups within the clustered ESTs. By assigning ESTs, cluster, gene family and domain information predictions on the possible function of the genes from which the ESTs derive can be made. In addition the physical maps of different species can be annotated using a common identifier (the gene family number) rather than the EST accession number. This unification of gene and EST nomenclature within and between species facilitates the analysis and comparison of genome maps. This information is publicly available from the Mendel-GFdb and Mendel-ESTS databases (www.mendel.ac.uk).