Workshop in the MIDST of Evaluating Drug Binding in Protein Variants Using Free Energy Perturbation
Sabanci University - Altunizade Campus, Istanbul, Turkey
Date: March 28, 2025
Computing the binding energy of a drug against a mutation can be challenging and prone to error. Moreover, many models (docking [1], FoldX [2], Rosetta [3], etc.) and deep learning tools (AlphaFold3 [4], DeepDTA [5], etc.) yield protein–drug complexes and energies with limited precision and explainability. Here, we employ free energy perturbation (FEP), a simulation method in which mutational perturbations are applied stepwise to an amino acid in a protein structure by annihilating the wild-type residue and creating the mutation—hence the term “alchemistry.” Because both the initial and final conformations are physical, this method enables the calculation of mutation and relative binding energies with high precision.
In this workshop, we use the enzyme dihydrofolate reductase (DHFR), a model protein for antibiotic resistance studies as an example system. We will perform FEP simulations on single mutational variants of DHFR bound to two drugs, TMP and 4'DTMP, and calculate the relative binding energies to determine the more favorable drug. This workshop will combine theory and application. In the morning, we will first learn about the fundamentals of FEP simulation and how to perform these simulations on a high-performance computer (TRUBA), after which we will run an FEP simulation on TRUBA. Each attendee will also run an FEP simulation using their temporary account. In the afternoon, we will discuss thermodynamic cycles, how to calculate relative binding energies, and how to analyze the FEP simulations. We will examine the outputs in depth through the lenses of computation, biophysics, and evolution.
Related DHFR references:
• https://doi.org/10.1093/molbev/msz086
• https://doi.org/10.1021/acs.jcim.3c00818
• https://doi.org/10.1002/cpe.8371
Level: Participants are expected to have minimal experience with MD simulations of proteins or similar systems
