Premium
A 1.8 Å resolution potential function for protein folding
Author(s) -
Crippen Gordon M.,
Snow Mark E.
Publication year - 1990
Publication title -
biopolymers
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.556
H-Index - 125
eISSN - 1097-0282
pISSN - 0006-3525
DOI - 10.1002/bip.360291014
Subject(s) - maxima and minima , parameterized complexity , globular protein , chemistry , protein folding , protein tertiary structure , folding (dsp implementation) , amino acid residue , function (biology) , ranging , resolution (logic) , crystallography , algorithm , peptide sequence , mathematics , computer science , mathematical analysis , artificial intelligence , telecommunications , biochemistry , evolutionary biology , biology , electrical engineering , gene , engineering
A general method is presented for constructing a potential function for approximate conformational calculations on globular proteins. The method involves solving a nonlinear program that seeks to adjust the potential's parameters in such a way that a minimum near the native remains a minimum and does not move far away, while any alternative minima shift so as to remain local minima but eventually rise higher than the level of the near‐native minimum. Although the potential trades computational speed for detail by representing each amino acid residue as only a single point, correct secondary structural preferences and reasonable tertiary folding can be built into the potential in an entirely routine way. The potential has been parameterized to agree with the crystal structure of avian pancreatic polypeptide (having 36 residues) in the sense that the lowest minimum found (‐407 arbitrary units) is reasonably close to the native (1.8 Å rms interresidue distance deviation). In contrast, the lowest nonnative conformation found after extensive searches by a variety of methods was −399 units and 7.5 Å away. Such potentials may prove to be useful in predicting approximate tertiary structure from amino acid sequence, if they can be generalized to apply to more than one protein.