A Novel Method for Force-Field Calibration Based on Maximum-Likelihood Approach and Thermal Unfolding Data

Abstract

Atomistic and coarse-grained simulations can be a great help in uncovering the mechanisms of physical processes at microscopic and mesoscopic levels at time scales ranging from femtoseconds to milliseconds. Any simulation study involves (1) setting up an appropriate simulation system representing the physical problem, (2) running the simulation and collecting information about the system, and (3) analyzing the collected data. The last step eventually leads to final conclusions about the system. Software for molecular simulation has been in development for many years and a number of high quality freely distributed general purpose simulation packages is available for researchers. Data analysis tools are usually less general as they often depend on a specific research project and the system under investigation. While many simulation packages come with a set of some general data analysis utilities, it is not unusual for such analysis tools to be developed on a per project basis inside research groups. Interestingly, there is a very limited set of available tools for setting up simulation systems, even though this is the very first and vital step of every simulation study. This lack of convenient general simulation system generators sometimes may even dictate the kind of simulations done based on the available initial systems rather than on the system being the best for a particular problem. In this work we describe a general software tool, bio.b-gen, for the creation of initial systems for biological molecular simulations. A number of case systems are demonstrated using an atomistic force field as well as the coarse grained MARTINI force field. The tool is designed to generate initial systems for the GROMACS general simulation package.