Maximum likelihood calibration of the UNRES force field for simulation of protein structure and dynamics
Paweł Krupa , Anna Hałabis , Wioletta Żmudzińska , Stanisław Ołdziej , H. A. Scheraga , Józef Adam Liwo
AbstractBy using the maximum likelihood method for force-field calibration recently developed in our laboratory, which is aimed at achieving the agreement between the simulated conformational ensembles of selected training proteins and the corresponding ensembles determined experimentally at various temperatures, the physics-based coarse-grained UNRES force field for simulations of protein structure and dynamics was optimized with seven small training proteins exhibiting a variety of secondary and tertiary structures. Four runs of optimization, in which the number of optimized force-field parameters was gradually increased, were carried out, and the resulting force fields were subsequently tested with a set of 22 α-, 12 β-, and 12 α + β-proteins not used in optimization. The variant in which energy-term weights, local, and correlation potentials, side-chain radii, and anisotropies were optimized turned out to be the most transferable and outperformed all previous versions of UNRES on the test set.
|Other language title versions|
|Journal series||Journal of Chemical Information and Modeling, ISSN 1549-9596|
|Publication size in sheets||0.65|
|Score|| = 40.0, 20-12-2017, ArticleFromJournal|
= 40.0, 20-12-2017, ArticleFromJournal
|Publication indicators||: 2016 = 3.76 (2) - 2016=4.066 (5)|
|Citation count*||2 (2018-04-24)|
* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.