Research

Research Policy

The recent development of measurement and data processing technologies has had a tremendous impact on the medical and healthcare fields, resulting in the accumulation of large-scale and diverse medical big data, including genomic and other omics data and real-world data such as receipts and electronic medical records. While the amount of medical big data is increasing, there is a need to develop new knowledge and new treatment methods that make use of this data.

The “Public Health Informatics Unit” is analyzing medical big data from various perspectives using data science, biostatistics, bioinformatics, epidemiology, and public health to 1) search for risk factors that lead to disease onset and severity, 2) develop algorithms to predict disease onset and severity, and 3) develop new treatment methods and new knowledge using the data. Based on these, we aim to advance the fields of medicine and health science by providing optimal preventive measures and interventions.

Through our research activities, we are also training data scientists in the fields of medicine and health science who can handle big data analysis.

Research Summary

The “Public Health Informatics Unit” conducts research utilizing large-scale medical and health data (Big Data), including genome information, receipts and electronic medical records, to search for risk factors leading to disease onset and severity, and to develop prediction algorithms for disease onset and severity.

Specific research themes include the following medical and health big data analysis research.

  1. Genomic data analysis of diseases and constitutions
  2. Development of disease diagnostic and disease onset prediction models utilizing multi-omics data obtained from blood
  3. Development of a prediction system for the onset of lifestyle-related diseases and mental disorders using big data from health checkups
  4. Prognosis prediction for newborns using maternal and neonatal data
  5. Search for risk factors that lead to certification of need for long-term care using health checkup data
  6. Other medical/health science projects

1. Analysis of genomic data on diseases and constitutions

Some people are born taller or shorter than others, some are stronger or weaker drinkers, some are more susceptible to disease than others, and there are individual differences in human constitutions. Genetic factors are thought to be the cause of these individual differences. In cooperation with research institutions in Japan and abroad, our laboratory is exploring genetic factors using genomic data such as single nucleotide polymorphisms (SNPs). So far, we have identified SNPs that contribute to the development of lifestyle-related diseases such as obesity, hypertension, and diabetes, as well as pancreatic cancer and other cancers. We are also developing models for predicting the risk of disease onset and disease progression using this genetic information, with the aim of realizing tailor-made medicine based on genome information.

 

<Examples of publications>

 Koyanagi Y, Nakatochi M et al. Genetic architecture of alcohol consumption identified by a genotype-stratified GWAS, and impact on esophageal cancer risk in Japanese. Science Advances 10(4):eade2780 (2024), DOI:10.1126/sciadv.ade2780, Link, Press release

 Lin Y, Nakatochi M, Hosono Y, Ito H et al. Genome-wide association meta-analysis identifies GP2 gene risk variants for pancreatic cancer. Nature Communications 11: 3175 (2020), DOI:10.1038/s41467-020-16711-w, Link

 Nakatochi M et al. Genome-wide meta-analysis identifies multiple novel loci associated with serum uric acid levels in Japanese individuals. Communications Biology 2:115 (2019), DOI: 10.1038/s42003-019-0339-0, Link

2. Development of disease diagnostic and disease onset prediction models utilizing multi-omics data obtained from blood

Recent developments in measurement technology have made it possible to obtain a comprehensive range of biological information. Such data are called omics data. In our laboratory, various omics data are obtained from the blood of local residents, and we are working on the identification of factors that enable disease diagnosis and prediction of the onset of lifestyle-related diseases, as well as the development of prediction models.

   ①DNA methylation(Epigenome)、②miRNA(Transcriptome)、③Protein(Proteome)  

 

<Examples of publications>

 Nakatochi M et al. Epigenome-wide association of myocardial infarction with DNA methylation sites at loci related to cardiovascular disease. Clinical Epigenetics 9:54 (2017), DOI: 10.1186/s13148-017-0353-3, Link

3. Development of a prediction system for the onset of lifestyle-related diseases and mental disorders using big data from health checkups

It is considered desirable to intervene appropriately in the development of lifestyle-related and psychiatric diseases from the stage when signs of disease onset appear, rather than starting treatment after the onset of the disease. Therefore, we are searching for risk factors for the onset of these diseases and developing a prediction model to detect signs of disease onset, using health examination data and genome data accumulated daily for residents in the Iwaki area of Hirosaki City, Aomori Prefecture.

 

Project website(Link):https://coi.hirosaki-u.ac.jp/

 

<Examples of publications>

Kinoshita F et al. Lifestyle parameters of Japanese agricultural and non-agricultural workers aged 60 years or older and less than 60 years: A cross-sectional observational study. PLoS ONE 18(10): e0290662 (2023), DOI: 10.1371/journal.pone.0290662, Link

4. Prognosis prediction for newborns using maternal and neonatal data

Active management of preterm infants has been shown to improve mortality rates. To this end, accurate prenatal prediction of mortality risk of preterm infants and identification of high-risk individuals is essential for early neonatal care. based on data from over 30,000 Japanese newborns and their mothers, we are developing algorithms to identify high-risk preterm infants using machine learning and AI.

5. Exploration for risk factors that lead to certification of need for care using certification data

In Japan, a long-term care insurance system was introduced in 2000 to provide benefits for the cost of long-term care services to those in need of care. Japan has a markedly aging population, and the number of people certified under these systems is increasing every year. The increase in the number of people certified as requiring support or nursing care causes various problems, such as financial pressure, increases in nursing care insurance premiums, and a rise in the ratio of admissions to nursing care facilities. Therefore, there is a need to extend healthy life expectancy through preventive nursing care interventions. To this end, it is important to provide appropriate interventions before people are certified as requiring long-term care, and it is necessary to identify those at high risk of needing long-term care at an early stage. Our laboratory is developing a prediction model for high-risk individuals by using statistical analysis techniques, machine learning, and AI to search for risk factors for the certification of persons requiring long-term care, based on information on the certification of persons requiring long-term care and health checkup data received from several municipalities in Aichi Prefecture.

 

<Examples of publications>

 Nakatochi M et al. U-shaped Link of Health Checkup Data and Need for Care using a Time-dependent Cox Regression Model with a Restricted Cubic Spline. Scientific Reports 13: 7537 (2023), DOI: 10.1038/s41598-023-33865-x, Link, Press release 

6. Others

We also encourage medical/health science assignments that students would like to do themselves with the big data that is currently available to the public or that they can acquire themselves.