Research on parallel algorithms for large scale Bayesian network structure search

• Project/Area Number: 15K20965
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Computational science; Intelligent informatics
• Research Institution: The University of Tokyo
• Principal Investigator: Tamada Yoshinori 東京大学, 医科学研究所, 特任講師 (80435495)
• Project Period (FY): 2015-04-01 – 2017-03-31
• Project Status: Completed (Fiscal Year 2016)
• Budget Amount: ¥2,340,000 (Direct Cost: ¥1,800,000、Indirect Cost: ¥540,000); Fiscal Year 2016: ¥910,000 (Direct Cost: ¥700,000、Indirect Cost: ¥210,000); Fiscal Year 2015: ¥1,430,000 (Direct Cost: ¥1,100,000、Indirect Cost: ¥330,000)
• Keywords: ベイジアンネットワーク / グラフ探索アルゴリズム / 並列アルゴリズム / 最適DAG探索アルゴリズム / 大規模並列計算

Outline of Final Research Achievements

We have successfully developed the novel parallel algorithm for searching for the optimal DAG structure using the supercomputer with 2D torus interconnect network. Using this algorithm we have succeeded in searching for the optimal DAG structure of the 36 variables and 200 samples using FX10 4800 computation nodes. This is the first example of searching for the optimal DAG structure with 36 variables in the literature. We have then improved this algorithm and succeeded in the optimal 37 variable DAG search using K computer 20768 computation nodes. Currently we are preparing the submission of the paper of this result.

Research Products

• [Journal Article] Efficient Parallel Algorithm for Optimal DAG Structure Search on Parallel Computer with Torus Network (2016)
  • Author(s): H. Honda, Y. Tamada, R. Suda
  • Journal Title: Lecture Notes in Computer Science Volume: 10048 Pages: 483-502
  • DOI: 10.1007/978-3-319-49583-5_37
  • ISBN: 9783319495828, 9783319495835
  • Peer Reviewed / Acknowledgement Compliant
• [Journal Article] 「京」を使った大規模データ解析によるがんのシステム異常の網羅的解析 (2016)
  • Author(s): 宮野 悟, 伊東 聰, Heewon Park, 白石 友一, 島村 徹平, 玉田 嘉紀, 井元 清哉
  • Journal Title: 実験医学増刊 Volume: 34 Pages: 733-738
• [Journal Article] The hnRNP-Htt axis regulates necrotic cell death induced by transcriptional repression through impaired RNA splicing (2016)
  • Author(s): Y. Mao, T. Tamura, Y. Yuki, D. Abe, Y. Tamada, S. Imoto, H. Tanaka, H. Homma, K. Tagawa, S. Miyano and H. Okazawa
  • Journal Title: Cell Death and Disease Volume: 7 Issue: 4 Pages: e2207-e2207
  • DOI: 10.1038/cddis.2016.101
  • Peer Reviewed / Open Access
• [Journal Article] Elevated β-catenin pathway as a novel target for patients with resistance to EGF receptor targeting drugs (2015)
  • Author(s): Nakata A, Yoshida R, Yamaguchi R, Yamauchi M, Tamada Y, Fujita A, Shimamura T, Imoto S, Higuchi T, Nomura M, Kimura T, Nokihara H, Higashiyama M, Kondoh K, Nishihara H, Tojo A, Yano S, Miyano S, Gotoh N
  • Journal Title: Scientific Report Volume: 5 Issue: 1 Pages: 13076-13076
  • DOI: 10.1038/srep13076
  • NAID: 120005764206
  • Peer Reviewed / Open Access
• [Journal Article] Oncogenic roles of TOPK and MELK, and effective growth suppression by small molecular inhibitors in kidney cancer cells (2015)
  • Author(s): T. Kato, H. Inoue, S. Imoto, Y. Tamada, T. Miyamoto, Y. Matsuo, Y. Nakamura and J.H. Park
  • Journal Title: Oncotarget Volume: 7 Issue: 14 Pages: 17652-17664
  • DOI: 10.18632/oncotarget.7755
  • Peer Reviewed / Open Access / Int'l Joint Research
• [Presentation] Cancer Genome Research with Gene Network Analysis using Bayesian Networks (2015)
  • Author(s): Y. Tamada
  • Organizer: The second Workshop on Advanced Methodologies for Bayesian Networks
  • Place of Presentation: Hiyoshi Campus, Keio University, 神奈川県横浜市
  • Year and Date: 2015-11-18
  • Int'l Joint Research
• [Presentation] The Cancer Network Galaxy: A Database of Cancer Gene Networks Estimated by SiGN-BN with K (2015)
  • Author(s): Y. Tamada, G. Tremmel, A. Niida, R. Yamaguchi, S. Imoto, and S. Miyano
  • Organizer: Supercomputational Life Science 2015
  • Place of Presentation: Takeda Hall, The University of Tokyo, 東京都文京区
  • Year and Date: 2015-10-20
  • Int'l Joint Research
• [Remarks] SiGN-BN
  • URL: http://sign.hgc.jp/signbn/index.html