And they're off... Neomorphic wins "race" at bioinformatics conference Neomorphic (Berkeley, Calif.; www.neomorphic.com) announced that Genie, its computational gene modeling system, provided highly accurate predictions in a "blind" experiment at the recent Intelligent Systems for Molecular Biology (ISMB) (ismb99.gmd.de) conference in Heidelberg, Germany. The goal of the experiment was to help identify the most accurate tools for annotating the Drosophila genome, the sequence of which is due to be completed by year-end. An open invitation for the competition was extended to the bioinformatics community at large, 3 months prior to the conference. Twelve research teams from around the world took up the challenge to find genes in three megabases of the Drosophila genome. Participants' predictions could be computational, biological or a combined consensus approach. Participants were asked to predict genes and other genomic features in a sample Drosophila genomic region that has been studied intensively in the past and thought to have 222 genes. The anonymous submissions were analyzed by a group of scientists at the Berkeley Drosophila Genome Center and judged by a panel of scientific assessors. Neomorphic's Genie turned in one of best performances. The program was able to cover 97% of the known gene sequence and of the genes found, 65% were exact matches over the entire length of the gene. The next closest group turned in only 56% exact matches. Furthermore, expert Drosophila biologists, using a combination of biologically determined cDNAs and multiple gene prediction tools, identified a total of 222 genes in the genomic sequence. Compared to that Genie found 256 genes, 36 additional genes that were not identified previously. Neomorphic's Genie substantially increases a scientist's methodology for gene identification as well as sequencing and analysis. Neomorphic's high scores in the Heidelberg Drosophila experiment bodes well for the Human Genome Project. David Kulp, vice president of Research for Neomorphic said, "With the encouraging results from the competition, we have every confidence to believe our software can analyze the human genome more effectively to discover novel genes and annotate these genes precisely." Web Site © Copyright 1999 by Lionheart Publishing, Inc. All rights reserved. Lionheart Publishing, Inc. 2555 Cumberland Parkway, Suite 299, Atlanta, GA 30339 USA Phone: 770-431-0867 | Fax: 770-432-6969 E-mail: lpi@lionhrtpub.com Web: www.lionheartpub.com Web Design by Premier Web Designs E-mail: lionwebmaster@preweb.com |