Large scale FMO-MP2 calculations on a massively parallel-vector computer

• Published in: Chemical physics letters 2008-05, Vol.457 (4), p.396-403
• Main Authors: Mochizuki, Yuji, Yamashita, Katsumi, Murase, Tadashi, Nakano, Tatsuya, Fukuzawa, Kaori, Takematsu, Kazutomo, Watanabe, Hirofumi, Tanaka, Shigenori
• Format: Article
• Language: English
• ISSN: 0009-2614; 1873-4448
• DOI: 10.1016/j.cplett.2008.03.090

Graphic representation of influenza hemagglutinin (HA) antigen–antibody system, taken from the PDB dataset of 1EO8. HA is a crucial glycoprotein existing on the surface of influenza virus, and it plays a central role in antigen–antibody reactions. HA is composed of HA1 and HA2 domains (colored in green and red, respectively), while the Fab-fragment of antibody consists of two chains (yellow and purple). This target system contains total 921 residues, and the number of atoms and 6-31G basis functions are 14 086 and 78 390, respectively. The FMO-MP2 job on this target can be completed in only 53.4 min with 4096 vector-processors of the Earth Simulator. From this calculation, the important sites in molecular recognition between antigen and antibody are revealed by interaction energies. The fragment molecular orbital (FMO) calculations have been successfully applied to a variety of realistic biochemical problems, by using our original A binit-mp program. In these applications, the inclusion of electron correlation through the second-order Møller-Plesset perturbation (MP2) was demonstrated to be essential to obtain qualitatively correct descriptions. Recently, the FMO calculations in A binit-mp were tuned for a massively parallel-vector processing. A series of FMO-MP2/6-31G calculations were performed on the Earth Simulator by which up to 4096 vector processors are available. The largest FMO-MP2 computation was carried out for an influenza hemagglutinin antigen–antibody system consisting of 921 residues, which was completed within one hour with 4096 processors.