Learning to Fight Infection

Research

View list of research

小林先生（750_500）.jpg

Researchers at The University of Tokyo found that analyzing short amino acid segments from T-cells could improve machine learning algorithms that predict the infection history of the donor. Based on the performance comparison using multiple diseases and sample size, the resulting algorithm could outperform existing solutions on smaller datasets. This work may enable a wider range of rare diseases to be diagnosed by future immunological blood testing.

2022.07.20

Press Release

Learning to Fight Infection

Researchers at the Institute of Industrial Science at The University of Tokyo introduce a machine learning method to help predict past infection from receptor sequences of immune T-cells even when little data is available, which may help improve human health and our understanding of adaptive immunity

Related Researchers

KOBAYASHI Tetsuya J Professor

Quantitative Biology

Department of Informatics and Electronics
Center for Research on Engineering in Medicine and Biology (CREMeB)
Center for Dependable Socio-information Platform（DSiP）

Mathematical Informatics : Graduate School of Information Science and Technology
Electrical Engineering and Information Systems : Graduate School of Engineering