Accurate ligand-protein binding affinities plays a crucial role in the lead optimization of drug design. The CUDA-accelerated PMEMD module of new AMBER 18 features improved GPU performance and support for GPU accelerated thermodynamic integration that delivers over two orders of magnitude greater performance than a single CPU core for the calculation of ligand-protein binding affinities, with no statistically significant numerical differences.
By watching this webinar replay, you will learn how this development in drug discovery is:
- Achievable: Now researchers can perform TI/FEP/MBAR on one GPU at the speed of 50-100 ns/day for drug design of a typical kinase protein. With adequate computational resources, it is now possible to optimize a lead in days or weeks, instead of months.
- Affordable: Compared to a large-scale CPU cluster, GPU-equipped workstations are much cheaper and much power-efficient. The widely used, academic AMBER simulation package is much more affordable than other commercial packages.
- Accurate and Reproducible: By extending the sampling with GPU acceleration, drug binding affinities can be predicted to within 0.1 kcal/mol precision. Through other modules in the Amber software distribution, researchers may tune ligand force-field parameters to reach higher accuracy for their specific drug design problems.