OVERVIEW

This site provides computer programs developed by the Institute of Statistical Mathematics team in JST CREST medical big data applications project. We have been developing statistical and machine learning methods for efficient large-scale genomic data analyses to develop effective molecular diagnostics and genetic biomarkers for precision medicine.

Team member

  • Hisashi Noma (Principal Investigator), Department of Data Science
    E-mail: noma[at]ism.ac.jp
  • Shuhei Mano, Department of Mathematical Analysis and Statistical Inference
    E-mail: smano[at]ism.ac.jp
  • Masayuki Henmi, Department of Data Science
    E-mail: henmi[at]ism.ac.jp
  • Takahiro Otani, Risk Analysis Research Center
    E-mail: otani[at]ism.ac.jp
  • Shonosuke Sugasawa, Risk Analysis Research Center
    E-mail: sugasawa[at]ism.ac.jp

* replace [at] to @

ACHIEVEMENTS

Assessments of multiple testing strategies in GWAS

A simulation program to compare the effectiveness of multiple testing procedures in genome-wide association studies (GWASs).

Programs

multtest.zip

(Usage: please see README.txt and tutorial.html in the zip file.)

Papers & Presentations

  1. Otani, T., Noma, H., Nishino, J. & Matsui, S. (2017). Re-assessment of multiple testing strategies for more efficient genome-wide association studies. ISM Research Memorandum 1202. [link]
  2. Otani, T., Noma, H., Nishino, J. & Matsui, S. (2015). Re-evaluation of the multiple testing strategy for detecting disease-related genetic variants in genome-wide association studies. East Asia Regional Biometric Conference 2015, Fukuoka, JAPAN.

Power assessments of GWAS using hierarchical mixture models

A tool for assessing statistical power of published GWASs using semiparametric hierarchical mixture models.

Programs

assesspwr.zip

(Usage: please see README.txt and tutorial.html in the zip file.)

Papers & Presentations

  1. Otani, T., Noma, H., Nishino, J. & Matsui, S. (2017). Re-assessment of multiple testing strategies for more efficient genome-wide association studies. ISM Research Memorandum 1202. [link]
  2. Otani, T., Noma, H., Nishino, J. Matsui, K. & Matsui, S. (2016). A comparison of multiple testing methods for effective strategies in genome-wide association studies. XXVIIIth International Biometric Conference, Victoria, CANADA.

Efficient association tests for rare variants

An optimal, powerful statistical test for simultaneously testing multiple rare variant effects without restrictive parametric assumptions.

Programs

ACST.zip

(Usage: please see README.txt and tutorial_ACST.html in the zip file.)

Papers & Presentations

  1. Sugasawa, S, Noma, H., Otani, T., Nishino, J. & Matsui, S. (2017). An efficient and flexible test for rare variant effects. European Journal of Human Genetics 25(6): 752-757. doi: 10.1038/ejhg.2017.43. [PubMed]

Exploring predictive biomarkers using multidimensional hierarchical mixture models

A tool for exploring predictive biomarkers from GWAS data in a randomized clinical trial via multidimensional hierarchical mixture models and an optimal discovery procedure.

Programs

rctgwas.zip

(Usage: please see README.txt and example.R in the zip file.)

Papers & Presentations

  1. Otani, T. & Noma, H. (2017). Efficient gene-by-treatment interactions test to develop predictive biomarkers on genome-wide studies using multidimensional hierarchical mixture models. ASHG 2017 Annual Meeting, Orlando, Florida.

CREST, Japan Science and Technology Agency

Advanced Application Technologies to Boost Big Data Utilization for Multiple-Field Scientific Discovery and Social Problem Solving
(Research Supervisor: Yuzuru Tanaka)

Exploring etiologies, sub-classification, and risk prediction of diseases based on big-data analysis of clinical and whole omics data in medicine
(Research Director: Tatsuhiko Tsunoda)

The Institute of Statistical Mathematics