立体構造情報に基づくタンパク質間相互作用ネットワーク予測 Protein-Protein Interaction Network Prediction Based on Tertiary Structure Data

Protein–protein interactions (PPIs) are fundamental in the majority of cellular processes and their study is of enormous biotechnological and therapeutic interest. The computational prediction for elucidation of PPI networks is crucial in biological fields. However, the development of an effective method to conduct exhaustive PPI screening represents a computational challenge.In this dissertation, we proposed a novel PPI network prediction system called MEGADOCK based on protein–protein docking calculation with protein tertiary structure information. MEGADOCK reduced the calculation time required for docking by using new score functions, rPSC and RDE, and was implemented on recent parallel high-performance computing environments by employing a hybrid parallelization with MPI and OpenMP and general-purpose graphics processing unit technique.We showed that MEGADOCK is capable of exhaustive PPI screening and completed docking calculations 9.8 times faster than the conventional method (Mintseris, et al. 2007) while maintaining an acceptable level of accuracy. When MEGADOCK was applied to a subset of a general benchmark dataset to predict 120 relevant interacting pairs from 14,400 protein combinations, an F-measure value of 0.231 was obtained. Moreover, the system was scalable as shown by measurements carried out on two supercomputing environments, TSUBAME 2.0 and K computer.It is now feasible to search and analyze PPIs while taking into account three-dimensional structures at the interactome scale. We demonstrated the applications to pathway analyses, bacterial chemotaxis, human apoptosis, and RNA binding proteins by using our system. As an example of the results, when analyzing the positive predictions of bacterial chemotaxis pathway from MEGADOCK, all the core signaling interactions were correctly predicted with the exception of interactions activated by protein phosphorylation.Large-scale PPI prediction using tertiary structures is an effective approach that has a wide range of potential applications. This method is especially useful for identifying novel PPIs of new pathways that control cellular behavior.

identifier:oai:t2r2.star.titech.ac.jp:50230937

Protein–protein interactions (PPIs) are fundamental in the majority of cellular processes and their study is of enormous biotechnological and therapeutic interest. The computational prediction for elucidation of PPI networks is crucial in biological fields. However, the development of an effective method to conduct exhaustive PPI screening represents a computational challenge. In this dissertation, we proposed a novel PPI network prediction system called MEGADOCK based on protein–protein docking calculation with protein tertiary structure information. MEGADOCK reduced the calculation time required for docking by using new score functions, rPSC and RDE, and was implemented on recent parallel high-performance computing environments by employing a hybrid parallelization with MPI and OpenMP and general-purpose graphics processing unit technique. We showed that MEGADOCK is capable of exhaustive PPI screening and completed docking calculations 9.8 times faster than the conventional method (Mintseris, et al. 2007) while maintaining an acceptable level of accuracy. When MEGADOCK was applied to a subset of a general benchmark dataset to predict 120 relevant interacting pairs from 14,400 protein combinations, an F-measure value of 0.231 was obtained. Moreover, the system was scalable as shown by measurements carried out on two supercomputing environments, TSUBAME 2.0 and K computer. It is now feasible to search and analyze PPIs while taking into account three-dimensional structures at the interactome scale. We demonstrated the applications to pathway analyses, bacterial chemotaxis, human apoptosis, and RNA binding proteins by using our system. As an example of the results, when analyzing the positive predictions of bacterial chemotaxis pathway from MEGADOCK, all the core signaling interactions were correctly predicted with the exception of interactions activated by protein phosphorylation. Large-scale PPI prediction using tertiary structures is an effective approach that has a wide range of potential applications. This method is especially useful for identifying novel PPIs of new pathways that control cellular behavior.

identifier:oai:t2r2.star.titech.ac.jp:71138020