Gene Family Identification Using Network Networks

摘要

With the exponential accumulation of sequence data, continued progress in the Human Genome Project will depend increasingly on advanced computational tools to manage and analyze the data. Utilizing information embedded within families of homologous sequences, a gene family identification approach may facilitate the understanding of gene functions. We have developed a GeneFIND (Gene Family Identification Network Design) system for database searching against gene families. It provides rapid and accurate protein clasification by combining several search/alignment tools, including our MOTIFIND neural networks and ProClass protein family database. GeneFIND has been implemented as a fullscale system and used for protein database organization. In this paper, we illustrate its effectiveness for genomic sequence analysis using Mycobacterium tuberculosis as an example. To further enhance oru neural network program, we also perform a comparative study of several neural network systems we have developed over the years. In the case study for signal peptide discrimination, back-propagation and counter-propagation neural networks, as well as various sequence encoding methods are analyzed. The resutls indicate that method which extracts both information contents and signals is most promising.