Fast protein fold recognition via sequence to structure alignment and contact capacity potentials.

摘要: We propose new empirical scoring potentials and associated alignment procedures for optimally aligning protein sequences to structures. The method has two main applications: first, the recognition of a plausible fold sequence unknown structure out database representative structures and, second, improvement alignments by using structural information in order find better starting point homology based modelling. function is derived from an analysis nonredundant known converting relative frequencies into pseudoenergies normalization according inverse Bolzmann law. These-so called contact capacity-potentials turn be discriminative enough detect folds absence significant similarity at same time simple allow very fast optimization procedure.