Gene Prediction by Pattern Recognition and Homology Search

This paper presents an algorithm for combining pattern recognition-based exon prediction and database homologysearch in gene model construction. Thegoal is to use homologousgenesor partial genesexisting in the databaseas reference modelswhile constructing (multiple) gene models fromexon candidates predicted by pattern recognition methods. A unified frameworkfor gene modelingis used for genes ranging from situations with strong homologyto no homologyin the database. To maximallyuse the homologyinformationavailable, the algorithm applies homology on three levels: (1) exon candidate evaluation, (2) gene-segmentconstruction with a reference model, and (3) (complete) gene modeling. liminary testing has been doneon the algorithm. Test results showthat (a) perfect genemodeling can be expectedwhenthe initial exonpredictions are reasonably good and a strong homologyexists in the database; (b) homology(not necessarily strong) in general helps improvethe accuracy of gene modeling;(c) multiple ge...

• Type: Conference
• Year: 1996
• Where: ISMB
• Authors: Ying Xu, Edward C. Uberbacher
• Keywords: Computational Biology, Gene, Gene Modeling, Gene Models, ISMB 1996