J. D. Chodera

Benchmarking Cross-Docking Strategies for Structure-Informed Machine Learning in Kinase Drug Discovery

This study benchmarks docking strategies for generating kinase:inhibitor complex geometries to improve machine learning scoring in drug discovery, finding that combining shape and electrostatics-based methods achieves the highest success rate.

Identifying and overcoming the sampling challenges in relative binding free energy calculations of a model protein:protein complex

This study identifies sampling challenges in protein:protein binding free energy calculations and demonstrates how to overcome them using alchemical methods and long simulations.

Death by a Thousand Cuts -- Combining Kinase Inhibitors for Selective Target Inhibition and Rational Polypharmacology

This paper presents a method to improve kinase inhibitor selectivity by combining inhibitors with divergent off-target activities, validated using a NanoBRET assay.

Capturing non-local through-bond effects in molecular mechanics force fields: II. Using fractional bond orders to fit torsion parameters

Turning high-throughput structural biology into predictive inhibitor design

This paper develops a machine learning framework to predict protein-ligand affinity from structural data, enabling the design of potent inhibitors against SARS-CoV-2 Main Protease.

Teaching free energy calculations to learn from experimental data

This paper uses quantum machine learning potentials to improve alchemical free energy calculations by overcoming limitations in force fields and enabling learning from experimental data.

Mutation in Abl kinase with altered drug binding kinetics indicates a novel mechanism of imatinib resistance

The paper identifies a mutation in Abl kinase that causes imatinib resistance through altered drug dissociation kinetics rather than reduced binding affinity.

Antibodies to the SARS-CoV-2 receptor-binding domain that maximize breadth and resistance to viral escape

This study identifies antibodies against SARS-CoV-2 that balance neutralization potency and breadth, with some showing high resistance to viral escape.

What Markov state models can and cannot do: Correlation versus path-based observables in protein folding models

This study evaluates the ability of Markov state models (MSMs) to reproduce path-based observables in protein folding, finding they are reliable only when state lifetimes exceed the lag time, and history-augmented MSMs improve reliability.

Open Science Discovery of Potent Non-Covalent SARS-CoV-2 Main Protease Inhibitors

The paper describes an open-science drug discovery campaign that identified a non-covalent, non-peptidic inhibitor of SARS-CoV-2 main protease using crowdsourcing and computational methods.

Fitting quantum machine learning potentials to experimental free energy data: Predicting tautomer ratios in solution

This paper uses quantum machine learning to predict tautomer ratios in solution by correcting limitations in quantum chemical thermochemistry approximations.

Overview of the SAMPL6 pKa Challenge: Evaluating small molecule microscopic and macroscopic pKa predictions

The paper evaluates pKa prediction methods for drug-like molecules, finding that top submissions achieved RMSE of 0.7-1.0 pKa units, but significant inaccuracies remain, especially for sulfur-containing heterocycles and multiprotic molecules.

Capturing non-local through-bond effects in molecular mechanics force fields I: Fragmenting molecules for quantum chemical torsion scans

The paper introduces a method to assess and improve fragmentation schemes for quantum chemical torsion scans by using fractional Wiberg bond orders to preserve chemical environment accuracy.

Towards chemical accuracy for alchemical free energy calculations with hybrid physics-based machine learning / molecular mechanics potentials

This paper improves the accuracy of alchemical free energy calculations for protein-ligand binding by combining machine learning and molecular mechanics potentials, reducing errors from 0.97 to 0.47 kcal mol-1.

Is structure based drug design ready for selectivity optimization?

This study evaluates the accuracy of predicting compound selectivity using free energy calculations for kinase inhibitors, finding that systematic errors correlate between targets and longer simulations improve selectivity prediction.

SARS-CoV-2 Simulations Go Exascale to Capture Spike Opening and Reveal Cryptic Pockets Across the Proteome

The paper uses exascale simulations to study SARS-CoV-2 protein dynamics, revealing cryptic epitopes and conformational changes that inform antiviral target design.

Assessing the accuracy of octanol-water partition coefficient predictions in the SAMPL6 Part II log P Challenge

This paper evaluates the accuracy of octanol-water partition coefficient predictions in the SAMPL6 log P Challenge, finding improved performance compared to previous challenges and highlighting the effectiveness of various methods.

The SAMPL6 SAMPLing challenge: Assessing the reliability and efficiency of binding free energy calculations

This study evaluates binding free energy calculation methods using the SAMPL6 challenge, finding variations in efficiency and reproducibility across different systems and methods.

Octanol-water partition coefficient measurements for the SAMPL6 Blind Prediction Challenge

This paper presents a dataset of octanol-water log P values for 11 compounds from the SAMPL6 Part II Partition Coefficient Challenge and describes the experimental protocol used to measure them.

Small-molecule targeting of MUSASHI RNA-binding activity in acute myeloid leukemia

This study identifies Ro 08-2750 as a small molecule that targets MSI2 RNA-binding activity, inducing differentiation, apoptosis, and reduced disease burden in AML models.

The Dynamic Conformational Landscapes of the Protein Methyltransferase SETD8

This study reveals dynamic conformational landscapes of SETD8 using X-ray structures and molecular simulations, showing how protein dynamics influence catalysis and cancer-associated mutations.

Overview of the SAMPL6 host-guest binding affinity prediction challenge

This paper provides an overview of the SAMPL6 host-guest binding affinity prediction challenge, highlighting methodological improvements and remaining challenges in predicting binding affinities.

pKa measurements for the SAMPL6 prediction challenge for a set of kinase inhibitor-like fragments

The paper evaluates pKa prediction methods using a dataset of kinase inhibitor-like fragments measured by UV absorbance and NMR.

OpenPathSampling: A Python framework for path sampling simulations. I. Basics

The paper introduces OpenPathSampling, a Python framework for path sampling simulations, providing tools for various transition path sampling methods and demonstrating their application to benchmark problems.

Bayesian analysis of isothermal titration calorimetry for binding thermodynamics

This paper introduces a Bayesian framework for analyzing ITC data to better estimate thermodynamic parameters and their uncertainties.

Predicting resistance of clinical Abl mutations to targeted kinase inhibitors using alchemical free-energy calculations

This paper uses alchemical free-energy calculations to predict how Abl mutations affect inhibitor affinity, achieving accurate results for clinical mutations.

Biomolecular simulations under realistic macroscopic salt conditions

This paper introduces a Monte Carlo osmostat to simulate biomolecules under dynamic salt conditions, showing significant differences in salt concentration distributions compared to fixed-salt simulations.

A dynamic mechanism for allosteric activation of Aurora kinase A by activation loop phosphorylation

This study shows that phosphorylation of Aurora A's activation loop activates the kinase without shifting the DFG-In state population, and that a DFG-Out subpopulation is important for optimized substrate binding kinetics.

OpenMM 7: Rapid Development of High Performance Algorithms for Molecular Dynamics

OpenMM 7 enables rapid development of high-performance molecular dynamics algorithms with extensible features that work across hardware types without modifying the core toolkit.

Approaches for calculating solvation free energies and enthalpies demonstrated with an update of the FreeSolv database

The paper updates the FreeSolv database with calculated solvation free energies, enthalpies, and entropies using a single protocol and provides input files for molecular simulations.