A Modular Learning Environment for Protein Modeling

Wepropose in this paper a modularlearning environmentfor proteinmodeling.In this system,the protein modelingproblemis tackledin twosuccessive phases. First, partial structural informationsare determinedvia numericallearningtechniques.Then,in the secondphase,the multipleavailableinformations are combinedin paaemmatchingsearchesvia dynamic programming.It is shownonreal problemsthat various protein structure predictionscan beimprovedin this way,suchas secondarystructure prediction,alignment of weaklyhomologousprotein sequencesor protein modelevaluations.

• Type: Conference
• Year: 1993
• Where: ISMB
• Authors: Jérôme Gracy, Laurent Chiche, Jean Sallantin
• Keywords: Computational Biology, ISMB 1993, Protein, System, The Protein Modelingproblemis