Implementation of massive parallel Evolutionaly Computation from Biological experiment data analysis obtained by wetGA

2017 Fiscal Year Final Research Report

• Project/Area Number: 26280095
• Research Category: Grant-in-Aid for Scientific Research (B)
• Allocation Type: Partial Multi-year Fund
• Section: 一般
• Research Field: Soft computing
• Research Institution: Tokyo Institute of Technology
• Principal Investigator: Yamamura Masayuki 東京工業大学, 情報理工学院, 教授 (00220442)
• Co-Investigator(Renkei-kenkyūsha): SAKAMOTO Kensaku 独立行政法人理化学研究所, 生命機能科学研究センター・非天然アミノ酸技術研究チーム, チームリーダー (50240685) ; SOMEYA Hiroshi 東海大学, 情報理工学部, 准教授 (00333518) ; Komiya Ken 東京工業大学, 情報理工学院, 助教 (20396790)
• Project Period (FY): 2014-04-01 – 2018-03-31
• Keywords: ウェットGA / 超並列進化計算 / WetTDGA / 分子進化 / NK地形 / 疑似適応度 / パラメータ最適化 / 細胞シミュレーション

Outline of Final Research Achievements

Evolutionary Computation is an optimization search technique inspired by Biological Evolution. We inversely developed a molecular implementation of Evolutionary Computation, called wet GA, and applied it in Protein Engineering. This research shows the next stage of this Biology / Computer spiral. We improved conventional Evolutionary Computation models with the knowledge obtained by wet GA. We found the fact that can be hard to achieve expected performance even with massive parallelism. There is a trade-off between the population size, i.e. the degree of parallelism, and the generation number, i.e. the time to find the optimum. We examined the trade-off by using travelling salesman problems and typical function optimization problems. The results show that too much parallelism will become wasteful when the problem has a simple search landscape.

Free Research Field

DNAコンピューティング