作者: Markéta Paloncýová , Gabin Fabre , Russell H. DeVane , Patrick Trouillas , Karel Berka
DOI: 10.1021/CT500419B
关键词: Statistical physics 、 Bilayer 、 Force field (chemistry) 、 Molecular dynamics 、 Computational chemistry 、 Membrane 、 Complex system 、 Molecule 、 Chemistry
摘要: Studies of drug-membrane interactions witness an ever-growing interest, as penetration, accumulation, and positioning drugs play a crucial role in drug delivery metabolism human body. Molecular dynamics simulations complement nicely experimental measurements provide us with new insight into interactions; however, the quality theoretical data dramatically depends on force field used. We calculated free energy profiles 11 molecules through model dimyristoylphosphatidylcholine (DMPC) membrane bilayer using five fields, namely Berger, Slipids, CHARMM36, GAFFlipids, GROMOS 43A1-S3. For sake comparison, we also employed semicontinuous tool COSMOmic. High correlation was observed between partition coefficients (log K). Partition by all-atomic fields (Slipids, GAFFlipids) COSMOmic differed less than 0.75 log units from experiment Slipids emerged best performing field. This work provides following recommendations (i) for global, systematic high throughput thermodynamic evaluations (e.g., K) is choice due to low computational costs; (ii) studies hydrophilic CHARMM36 should be considered; (iii) more complex systems, taking account all pros cons, choice.