The increasing availability of human genetic data coupled with the steady decline in the cost of genomic data acquisition has made it possible to analyze large-scale data of hundreds of thousands of individuals. Specifically, it offers unique opportunities for breakthroughs in biomedical research, such as therapeutic target discovery and precise assessment of the genetic liability of diseases. The frontier in future genomics research is the methodological innovations in computational and statistical analyses.
Yosuke Tanigawa, a postdoctoral researcher at the Massachusetts Institute of Technology (MIT), investigates the genomic basis of diseases using techniques in statistical genetics and computational biology. Methodologically, he develops statistical models that jointly analyze the genetics of multiple disease outcomes and medically relevant traits. On the application front, he analyzes large-scale genetic datasets from hundreds of thousands of individuals. The combination of these parallel efforts brings insights into the genetic basis of the disease, enabling us to assess how subtle differences in personal genomic information influence disease liability and to nominate potential therapeutic targets.
Many diseases and medically-relevant traits are influenced by many genetic variants along with environmental factors. With polygenic risk score (PRS), a statistical approach that aggregates the effects of multiple genetic variants, one can estimate genetic predisposition on disease liability for individuals from their genetic information. Tanigawa co-led a large-scale genetic study, analyzing datasets from more than 360,000 individuals, and developed PRS models for blood and urine biomarkers. The research team also showed that a ‘multi-PRS’ approach, which combines a single-trait PRS model for a disease and 35 PRS models for biomarkers, improves the performance of genetics-based prediction of disease liability. For chronic kidney diseases, for example, they show that the conventional single-trait PRS model performs no better than random classifiers, whereas their multi-PRS approach yields improved and statistically significant performance in genetic-based disease liability prediction, highlighting the value of their methodological innovation.
On the application front, Tanigawa’s genetic research also contributes to drug discovery.
Tanigawa and his colleagues investigated the impacts of rare genetic variants on glaucoma, the leading cause of blindness in Japan and the second leading cause of blindness worldwide. Analyzing more than 500,000 individuals in the UK and Finland, the team identified an allelic series of rare genetic variants in the ANGPTL7 gene that reduce the risk of glaucoma by more than 30%. The study was selected as the Editor's Choice in Science journal and attracted significant attention. Pharmaceutical companies are pursuing research and development for therapeutics that target ANGPTL7.
“With large-scale genetic information, we have an increasing ability to draw mechanistic insights into disease onset, progression, and their differences in outcomes across individuals. My current research focus is on disease heterogeneity. I envision computational and statistical analysis of large-scale genomic data will bring new insights into biology and medicine,” said Tanigawa with great enthusiasm.