Adaptive lambda square dynamics simulation; An Efficient conformational sampling method for biomolecules

Ikebe, Kimiyoshi

Free energy landscapes of biomolecules provide valuable information on how the molecule behaves and obtaining them is one of the most important objectives in biology. Multicanonical molecular dynamics (MD) simulation is a powerful technique developed to efficiently investigate the landscape. However, we need still a more efficient method to obtain a reliable ensemble for a larger and more complicate biomolecule system. To this end, Okumura developed partial McMD that enhances the sampling only for partial energy term. The focused sampling will reduce the space to be sampled and increase the sampling efficiency. However, we found that partial McMD can have a less sampling efficiency than McMD in a system for a poly-lysine decapeptide. In this seminar, we address the fundamental reason and report that a new simulation method, adaptive lambda square dynamics (ALSD), which can resolve the problem occurring when using partial McMD, greatly increased the sampling efficiency over McMD.