Development of a new data cleaning method for questionnaires used in large cohorts

• Project/Area Number: 18K10099
• Research Category: Grant-in-Aid for Scientific Research (C)
• Allocation Type: Multi-year Fund
• Section: 一般
• Review Section: Basic Section 58030:Hygiene and public health-related: excluding laboratory approach
• Research Institution: Tohoku University
• Principal Investigator: MAKINO Satoshi 東北大学, 東北メディカル・メガバンク機構, 助教 (30423403)
• Co-Investigator(Kenkyū-buntansha): 田宮 元 東北大学, 東北メディカル・メガバンク機構, 教授 (10317745) ; 櫻井 利恵子 東北大学, 東北メディカル・メガバンク機構, 非常勤講師 (50794541)
• Project Period (FY): 2018-04-01 – 2021-03-31
• Project Status: Completed (Fiscal Year 2020)
• Budget Amount: ¥3,900,000 (Direct Cost: ¥3,000,000、Indirect Cost: ¥900,000); Fiscal Year 2020: ¥1,300,000 (Direct Cost: ¥1,000,000、Indirect Cost: ¥300,000); Fiscal Year 2019: ¥1,040,000 (Direct Cost: ¥800,000、Indirect Cost: ¥240,000); Fiscal Year 2018: ¥1,560,000 (Direct Cost: ¥1,200,000、Indirect Cost: ¥360,000)
• Keywords: コホート研究 / データクリーニング / 外れ値検出

Outline of Final Research Achievements

In large-scale genomic cohort studies, in spite of careful study planning and implementation, and the introduction of error prevention methods, various errors are inevitable, and these errors may have a significant impact on the study results. However, it was not possible to manually clean a large number of questionnaires. Therefore, by using a statistical model that extends principal component analysis (PCA) by utilizing known information when detecting outliers from the data population, we developed and implemented a method to automate the detection of candidate errors and to improve its accuracy.

Academic Significance and Societal Importance of the Research Achievements

データクリーニングは、大規模コホート研究のみならず、その重要性が認識されているものの、世界的にコンセンサスを得られた手法は存在しなかった。海外の大規模コホートにおいても、多くはタッチスクリーンベースであるためデータ入力時のエラー発生率は低いと考えられるものの、単純なミスマッチやデータ形式の違いを検出しているのみである。本研究はパターンの違いをエラー検出に利用するため、これまで事実上不可能であった調査票の経時的データや家族間のデータのクリーニングに関しても応用可能となった。

Research Products

• [Journal Article] Clustering by phenotype and genome-wide association study in autism (2020)
  • Author(s): Narita Akira, Nagai Masato, Mizuno Satoshi, Ogishima Soichi et al.
  • Journal Title: Translational Psychiatry Volume: 10 Issue: 1 Pages: 290-290
  • DOI: 10.1038/s41398-020-00951-x
  • Peer Reviewed / Open Access
• [Journal Article] Genome-wide association study identifies new loci for albuminuria in the Japanese population (2020)
  • Author(s): Hiroshi Okuda, Koji Okamoto, Michiaki Abe, Kota Ishizawa, Satoshi Makino, Osamu Tanabe, Junichi Sugawara, Atsushi Hozawa, Kozo Tanno, Makoto Sasaki, Gen Tamiya, Masayuki Yamamoto, Sadayoshi Ito, Tadashi Ishii
  • Journal Title: Clinical and Experimental Nephrology Volume: 印刷中 Issue: 8 Pages: 1-9
  • DOI: 10.1007/s10157-020-01884-x
  • Peer Reviewed / Open Access
• [Journal Article] 3.5KJPNv2: an allele frequency panel of 3552 Japanese individuals including the X chromosome (2019)
  • Author(s): Tadaka S. et al. Koshiba S. et al. Kinoshita K.（合計２５名、本研究者は１４番目）
  • Journal Title: Hum Genome Var Volume: 6 Issue: 1 Pages: 28-28
  • DOI: 10.1038/s41439-019-0059-5
  • Peer Reviewed / Open Access
• [Journal Article] Outlier detection for questionnaire data in biobanks (2019)
  • Author(s): Sakurai Rieko、Ueki Masao、Makino Satoshi、Hozawa Atsushi、Kuriyama Shinichi、Takai-Igarashi Takako、Kinoshita Kengo、Yamamoto Masayuki、Tamiya Gen
  • Journal Title: International Journal of Epidemiology Volume: 印刷中 Issue: 4 Pages: 1305-1315
  • DOI: 10.1093/ije/dyz012
  • Peer Reviewed
• [Journal Article] Goodness-of-fit test for the parametric proportional hazard regression model with interval-censored data (2018)
  • Author(s): Sakurai Rieko、Hattori Satoshi
  • Journal Title: Biostatistics & Epidemiology Volume: 2 Issue: 1 Pages: 115-131
  • DOI: 10.1080/24709360.2018.1529347
  • Peer Reviewed
• [Presentation] バイオバンクにおける質問票データに対する外れ値検出 (2020)
  • Author(s): 櫻井利恵子
  • Organizer: 第2回日本メディカルAI学会学術集会