FMODB: The world’s first database for protein quantum chemistry calculations
FMODD has developed the “FMO Database (FMODB)“ to archive FMO data of thousands of proteins. FMODB provides a simple interface to access and download FMO data via the Internet.
The archived FMO data include the fragment interaction energy (IFIE, also called PIE) and its energy decomposition (PIEDA). IFIE and PIEDA evaluate cation-π, CH/π, π-π, and halogen bond interactions, which cannot be precisely calculated with conventional classical molecular mechanical (MM) force fields. The abundant information on the interactions obtained by FMO is useful for researchers in various scientific fields. For non-specialists, however, it is difficult to prepare structures and parameters to start FMO calculation and to analyze the calculated results. Therefore, FMODB compiles data obtained by FMO so that anyone can access and use them easily and effectively.
Construction of AI using big data has recently been flourishing. Regarding this matter, we hope that energy values obtained by FMO, namely IFIE, PIE, and PIEDA, may serve as descriptors of statistical approaches for protein-ligand interaction analysis. Furthermore, we aim to construct a high-precision MM force field at the level of quantum chemistry by fitting MM parameters to physical properties obtained by FMO, such as partial atomic charges, with a help of AI.
FMO Database (FMODB)
C. Watanabe et al, (2019）Chem.-Bio. Inf. J. 19, 5-18. https://doi.org/10.1273/cbij.19.5
Excellent Outcome Award of “K computer” for two consecutive years
Both in FY2016 and FY2017, we won the outstanding achievement award given to remarkable researches performed on the K-computer. FMODD has been highly evaluated because of the effective use of the Japanese supercomputer based on the theory originated in Japan.
“Construction of FMO drug design platform utilizing HPCI” (Project IDs: hp160103, hp170183)