MonteGrappa: An iterative Monte Carlo program to optimize biomolecular potentials in simplified models

Abstract

Simplified models, including implicit-solvent and coarse-grained models, are useful tools to investigate the physical properties of biological macromolecules of large size, like protein complexes, large DNA/RNA strands and chromatin fibres. While advanced Monte Carlo techniques are quite efficient in sampling the conformational space of such models, the availability of realistic potentials is still a limitation to their general applicability. The recent development of a computational scheme capable of designing potentials to reproduce any kind of experimental data that can be expressed as thermal averages of conformational properties of the system has partially alleviated the problem. Here we present a program that implements the optimization of the potential with respect to the experimental data through an iterative Monte Carlo algorithm and a rescaling of the probability of the sampled conformations. The Monte Carlo sampling includes several types of moves, suitable for different kinds of system, and various sampling schemes, such as fixed-temperature, replica-exchange and adaptive simulated tempering. The conformational properties whose thermal averages are used as inputs currently include contact functions, distances and functions of distances, but can be easily extended to any function of the coordinates of the system. Program summaryProgram title: MonteGrappaCatalogue identifier: AEUO_v1_0Program summary URL:http://cpc.cs.qub.ac.uk/summaries/AEUO_v1_0.htmlProgram obtainable from: CPC Program Library, Queen’s University, Belfast, N. IrelandLicensing provisions: GNU General Public License, version 3No. of lines in distributed program, including test data, etc.: 139,987No. of bytes in distributed program, including test data, etc.: 1,889,541Distribution format: tar.gzProgramming language: C.Computer: Any computer with C compilers.Operating system: Linux, Unix, OSX.RAM: Bytes depend on the size of the system, typically 4 GBClassification: 3, 16.1.External routines: gsl, MPI (optional)Nature of problem:Optimize an interaction potential for coarse-grained models of biopolymers based on experimental data expressed as averages of conformational properties.Solution method:Iterative Monte Carlo sampling coupled with minimization of the chi2 between experimental and back-calculated data making use of a reweighting algorithm.Running time:Hours to days, depending on the complexity of the problem.