Collaborative Research: CDS&E: Theory and Methods for Implicit Molecular Solvation in Ligand and Ion Binding

Tyler Luchko of California State University, Northridge (CSUN) and David Case of Rutgers University are supported by an award from the Chemical Theory, Models and Computational Methods program in the Division of Chemistry to reduce the computational cost of calculating the binding of small molecules and ions to proteins, DNA and RNA. The ability of computer models to make accurate quantitative predictions about how effectively molecules bind with (or “stick” to) each other in a liquid environment is key to understanding the healthy functioning of biological systems or the structure and stability of advanced materials.

But the models required to achieve the desired accuracy often come with impractical computational costs. To address this, Drs. Luchko and Case will develop methods to improve the computational efficiency of predicting the tight binding of drug-like molecules to biological and non-biological targets and also the diffuse (“cloud-like”) binding of ions around DNA and RNA.

They will be pursuing new algorithms that offer a high degree of accuracy in the required calculations while also lowering the computational cost, by taking advantage of advancing hardware options, like GPUs, to speed up calculations. Both aspects will require iterative testing and tailoring to achieve maximum effectiveness. These methods and software will be freely available for broad application as part of the AmberTools molecular modeling suite.

This work will directly involve students at CSUN, of whom over 50% are from traditionally underserved groups, thereby contributing to the training of a diverse STEM workforce.

Central to the methods that will be developed is the 3D reference interaction site model (3D-RISM) of molecular liquids. 3D-RISM is a potentially powerful tool for molecular modeling, as it rapidly provides equilibrium solvent density distributions and accurate solvation free energies; however, its computational cost is too high for many molecular binding applications at present. To preserve the strengths of 3D-RISM while reducing the overall computational load, Drs.

Luchko and Case will develop methods to address the specific challenges of the tight binding of small, drug-like molecules and the diffuse binding of ions around DNA and RNA. For small, drug-like molecules, he and his group will create a thermodynamic cycle that employs fast implicit solvent methods to efficiently include sampling, followed by a 3D-RISM correction.

For diffuse ion binding, they will optimize ion force field parameters and solvers for 3D-RISM to predict the stability of DNA and RNA at different concentrations of divalent ions. In addition, they will implement 3D-RISM on GPUs, to directly reduce the computational cost of the model for all 3D-RISM for these and all other applications. Combined, these objectives will provide easy-to-use tools and solvent models for calculating binding free energies that do not sacrifice speed or accuracy.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.