September 2015 Summaries

Free energy perturbation (FEP) is a crucial simulation method in the life sciences industry for calculating the free energy difference between two chemical states, making it particularly advantageous in drug development due to its scalability in high-performance environments. This method involves alchemically transforming one chemical state into another through intermediate non-physical states, allowing for parallel simulations that enhance computational efficiency. By using platforms like Rescale, researchers can effectively run multiple independent simulations, each corresponding to specific lambda values, and potentially increase sampling efficiency through Hamiltonian Replica Exchange if supported by their software. An example from Alchemistry.org demonstrates how to set up a free energy calculation for ethanol using GROMACS on Rescale, detailing the necessary input files and simulation settings. The process benefits from optimal hardware usage by mapping threads to physical cores, ultimately leading to efficient performance metrics. The example highlights the ease of deploying such simulations on the Rescale platform and encourages researchers in the pharmaceutical sector to leverage these tools for advancing drug development, offering assistance for those unfamiliar with the process.

ITOCHU Techno-Solutions Corporation (CTC) has partnered with Rescale, Inc. to offer cloud-based high performance computing (HPC) platforms to Japanese enterprises, particularly in manufacturing, construction, and pharmaceuticals, to enhance their engineering and science simulations. Rescale's platform, supported by an extensive infrastructure of over 8 million servers across 30 global data centers, provides a secure and scalable HPC environment, integrating more than 100 simulation software tools suitable for a range of industries. This collaboration allows companies to access customizable HPC resources through a web-based interface or API, significantly reducing analysis times and enabling more efficient R&D processes. The pay-per-use model and advanced computing capabilities, including GPUs and InfiniBand technology, further optimize simulation efforts, allowing businesses to focus on core research activities by minimizing the time and cost associated with building internal HPC clusters.