Research and cloud deployment of enhanced sampling methods in MovableType

Abstract The study of protein/ligand binding is one of the central problems in computational biology because of its importance in understanding intermolecular interactions, and because of its practical payoff in drug discovery efforts. The transformative impact accurate target/ligand structure can have in the design of next generation

medicines cannot be overstated. If we could routinely and accurately design molecules using these approaches it would revolutionize drug discovery by winnowing out compounds with no activity while focusing more effort and scrutiny on highly active compounds. In this proposal we describe a novel method we call MovableType (MT) that for the first time will be coupled with

cutting edge enhanced molecular dynamics (MD) methods (e.g., Simulated Tempering, Accelerated MD, Metadynamics, and Replica exchange MD) in Aims I.1 and II.1a, linear scaling quantum mechanics (for improved electrostatics) in Aim I.2, and a new Monte Carlo sampling regime called Consecutive Histograms

Monte Carlo (CHMC) in Aim II.1b for increased speed. We expect this development to significantly expand the domain applicability of MT in particular (and free energy methods in general) to include those situations which require greater conformational sampling than can be provided by docking alone. MT addresses the protein ligand binding and scoring problem using fundamental statistical mechanics combined

with a new way to generate the ensemble of a ligand in a protein binding pocket. Via a rapid assembly of the necessary partition functions, with MT we directly obtain absolute binding free energies and the low free energy poses (versus most conventional free energy methods in commercial/industrial labs which usually obtain relative

binding free energies). Conceptually, the MT method is analogous to block and type set printing, which allows us to efficiently evaluate partition functions describing regions or systems of interest. Overall, the MT method is a general one and can use a broad range of two-body potential functions and can be extended to higher-order

interactions if so desired. Recent work with the MT method has led to the launch of three core product modules: MTScore (both end state and ensemble-based binding affinity prediction), MTDock (ligand placement), and MTCS (ligand conformational search). In this project, we will extend our MT product line by optimizing the method for

use with advanced sampling techniques and deliver this methodology to computational chemists for use in their industrial structure-based drug design campaigns. This work will involve development of a new, integrated tool for automated structure/model preparation, integration with and optimization for several molecular dynamics

engines, addition an updated electrostatics engine (built on our mature, linear scaling, semi-empirical quantum mechanics infrastructure), development of a new Monte Carlo method for increased speed, and cloud-based deployment on the GridMarkets platform (Aim II.2). Finally, in Aim II.3, we will commercially deploy the

technology, construct graphical user interfaces for use in MOE, and validate its use in real life structure-based drug discovery problems with our pharmaceutical collaborators (see Letters of Support).