Improvement of MEGADOCK-WEB integrated database for predicted protein-protein interactions, and its coordination with cloud environment for on-demand dockings
We developed a protein - protein docking software, called MEGADOCK, to understand protein - protein interactions (PPIs) comprehensively. We are also developing a database of predicted PPIs and its web - based interface, named MEGADOCK - WEB, to make users access a vast amount of PPIs predicted by MEGADOCK easily. In this study, we solved the problem about latency caused by the huge data size by reducing the number of accesses to other databases. Thereby, approximately 36 million predicted PPIs can be browsed on MEGADOCK - WEB. Furthermore, we implemented on MEGADOCK - WEB a new feature of on - demand PPI prediction between two input structures which are not registered in the datadase yet. In order to facilitate docking calculation by MEGADOCK as the back - end on cloud computing environment, we developed dedicated APIs. Via these APIs, MEGADOCK - WEB can work together with the back - end MEGADOCK on cloud computing environment.