AutoParams: An Automated Web-Based Tool To Generate Force Field Parameters for Molecular Dynamics Simulations

Area of Science:

• Computational chemistry
• Biophysics
• Materials science

Background:

• Molecular dynamics (MD) simulations are crucial for studying molecular motion and conformation over time.
• Accurate force fields are essential for MD simulations, but creating parameters for unusual molecules is challenging and time-consuming.
• The growing need for large datasets in machine learning and advancements in computational power highlight the importance of efficient parameter generation.

Purpose of the Study:

• To present AutoParams, an open-source automated service for generating Amber force field parameter sets.
• To address the limitations of standard force fields for complex and novel molecular structures.
• To streamline the creation of force field databases for enhanced molecular simulations.

Main Methods:

• AutoParams utilizes hierarchical atom-typing logic for broad applicability.
• The service accepts minimal user input to generate parameter sets for diverse molecules, including those combining different molecular types.
• It integrates with charge generation programs like Psi4, PsiRESP, and TeraChem and includes error checking and testing protocols.

Main Results:

• AutoParams successfully generates useful Amber force field parameters for a wide range of molecules, especially complex ones.
• The service automates a previously manual and error-prone process, saving significant user time.
• It facilitates the creation of comprehensive force field databases essential for large-scale simulations and machine learning.

Conclusions:

• AutoParams provides an efficient, automated solution for generating molecular dynamics force field parameters.
• The open-source tool enhances the accessibility and reliability of parameter creation for computational chemistry research.
• This advancement supports the development of more extensive and accurate molecular simulations and machine learning models.