Fast rational enzyme design by computational non-equilibrium alchemical transformations

Custom-designed enzymes offer potential for sustainable fine chemical production, but traditional experimental methods used for their design are often inefficient and labor-intensive. Here, we propose a computational workflow that predicts changes in activation free energy barriers caused by mutations. This approach uses non-equilibrium alchemical free energy calculation with <i>ab initio</i> derived force fields to predict how mutations affect the rate-limiting step in enzyme kinetics. We applied the methodology to two enzymes that catalyze the hydride transfer from NADPH to their respective substrates, achieving results closely matching experimental data with minimal errors of only a few kJ mol^-1. Additionally, its low computational requirements make it perfect for high-throughput analyses, aiding in rational enzyme design.