Automated Assessment of Binding Affinity via Alchemical Free Energy Calculations

Abstract

Free energy calculations have seen increased usage in structure-based drug design. Despite the rising interest, automation of the complex calculations and subsequent analysis of their results are still hampered by the restricted choice of available tools. In this work, an application for automated setup and processing of free energy calculations is presented. Several sanity checks for assessing the reliability of the calculations were implemented, constituting a distinct advantage over existing open-source tools. The underlying workflow is built on top of the software Sire, SOMD, BioSimSpace and OpenMM and uses the AMBER14SB and GAFF2.1 force fields. It was validated on two datasets originally composed by Schrödinger, consisting of 14 protein structures and 220 ligands. Predicted binding affinities were in good agreement with experimental values. For the larger dataset the average correlation coefficient Rp was 0.70 ± 0.05 and average Kendall’s τ was 0.53 ± 0.05 which is broadly comparable to or better than previously reported results using other methods. <br>

Sire: An Interoperability Engine for Prototyping Algorithms and Exchanging Information Between Molecular Simulation Programs

“…The somd application was developed by interfacing sire's IO, energy and computer algebra functionality with GPU-accelerated MD functionality of the OpenMM library to create a single topology alchemical molecular dynamics free energy calculation engine. somd has been used extensively over the years to develop alchemical free energy calculation methodologies, and to support drug discovery projects 6,12,30,[34][35][36][37][38][39][40][41][42][43][44][45][46][47][48] .…”

“…Commercially, the sire applications somd, waterswap and BioSimSpace have been embedded in Flare produced by Cresset [10][11][12] , whereas BioSimSpace is used within the drug discovery pipelines developed by Exscientia.ai 13 .…”

“…Software developed using sire includes waterswap [1][2][3] , quantomm 4 , nautilus 5 , somd 6 , FESetup 7 and most notably BioSimSpace 8,9 . Commercially, sire is embedded in Flare produced by Cresset [10][11][12] and within the drug discovery pipelines used by Exscientia.ai 13 .…”

SARS-CoV-2 and MERS-CoV spike protein binding studies support stable mimic of bound 9-O-acetylated sialic acids

“…Ligands were transformed over 13-21 linearly-spaced λ windows, based on a previous study that used SOMD and other alchemical free energy methods in an assessment of binding affinities 55 , and following the scheme outlined in the SI of Loeffler et al 54 In general, the first set of simulations for each transformation was ran with 13 windows, and the second set of simulations with 21 for greater certainty. Each simulation was minimized over 1000 steps prior to 5 ns production run in the NPT ensemble, where free energies were computed from the last 4 ns, based on the assessments from Kuhn et al and Mey et al 55,56 Each transformation was computed in the reverse direction, and simulations were repeated in the forward and reverse directions, starting from the same equilibrated starting complex. For example, Neu5,9Ac 2 was transformed to Neu5Ac9NAc by slowly turning off the interactions of Neu5,9Ac 2 and slowly turning them on for Neu5Ac9NAc, over a set of discrete steps, as described the SI of Loeffler et al 54 .…”

A Hybrid Alchemical Free Energy and Machine Learning Methodology for the Calculation of Absolute Hydration Free Energies of Small Molecules

“…[1][2][3][4] The domain of applicability of current alchemical methodologies has to date mainly been restricted to hit-to-lead and lead optimisation scenarios owing to limitations in computing cost, conformational sampling, and the accuracy of the potential energy functions used to compute protein-ligand energetics. [5][6][7][8][9][10] There is continued interest in the development of more accurate potential energy functions to benchmark FEP workflows on diverse well curated diverse protein-ligand datasets, [11][12][13][14] and for applications to blinded challenges or methodological studies. [15][16][17][18][19][20] The calculation of hydration free energies has historically been an important stepping stone towards more accurate forcefields for protein-ligand binding free energy calculations.…”