Introduction of a bounded penalty function in contact‐assisted simulations of protein structures to omit false restraints

Józef Adam Liwo , Emilia Lubecka

Abstract

Contact‐assisted simulations, the contacts being predicted or determined experimentally, have become very important in the determination of the structures of proteins and other biological macromolecules. In this work, the effect of contact‐distance restraints on the simulated structures was investigated with the use of multiplexed replica exchange simulations with the coarse‐grained UNRES force field. A modified bounded flat‐bottom restraint function that does not generate a gradient when a restraint cannot be satisfied was implemented. Calculations were run with (i) a set of four small proteins, with contact restraints derived from experimental structures, and (ii) selected CASP11 and CASP12 targets, with restraints as used at prediction time. The bounded penalty function largely omitted false contacts, which were usually inconsistent. It was found that at least 20% of correct contacts must be present in the restraint set to improve model quality with respect to unrestrained simulations.
Publication typeIn press (online first, early view)
Author Józef Adam Liwo (FCh / DTCh / LMM)
Józef Adam Liwo,,
- Laboratory of Molecular Modeling
, Emilia Lubecka (FMPI / II)
Emilia Lubecka,,
- Institute of Informatics
Journal seriesJournal of Computational Chemistry, ISSN 0192-8651, [1096-987X], (A 35 pkt)
Issue year2019
Noonline first
Pages1-15
Publication size in sheets0.7
Keywords in Englishrestrained simulations, bounded restrained function, multiplexed replica-exchange molecular dynamics, UNRES, coarse-graining
ASJC Classification2605 Computational Mathematics; 1600 General Chemistry
DOIDOI:10.1002/jcc.25847
URL https://doi.org/10.1002/jcc.25847
Languageen angielski
Score (nominal)35
ScoreMinisterial score = 35.0, 15-07-2019, ArticleFromJournal
Ministerial score (2013-2016) = 35.0, 15-07-2019, ArticleFromJournal
Publication indicators Scopus SNIP (Source Normalised Impact per Paper): 2016 = 1.297; WoS Impact Factor: 2017 = 3.221 (2) - 2017=4.161 (5)
Citation count*
Cite
Share Share

Get link to the record


* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.
Back