Tetsuya J. Kobayashi

This paper introduces an unsupervised method to detect low-frequency antigen-specific TCRs using distance-based anomaly scoring, achieving higher accuracy than existing methods in three immunological contexts.

The paper models mutation probability in Arabidopsis thaliana using epigenetic data, identifying combinatorial patterns that predict mutation-rate variation across coding sequences.

This paper introduces a deep learning model to predict gene regulation relationships for both known and unknown transcription factors, achieving higher accuracy for unknown TFs than existing methods.

The paper quantifies how irradiation affects iNKT cell development in the thymus of mice by analyzing TCR repertoire changes and recovery dynamics.

The paper presents a theoretical framework for cell death by modeling it as a controllability problem in dynamical systems, using stoichiometric rays to define life-death boundaries.

The paper develops a deep learning algorithm to predict mouse embryo live birth potential from nuclear structural features in fluorescence images, achieving 83.87% accuracy.

The paper uses neural networks to infer size control laws in bacteria and fission yeast from noisy time-series data, revealing known mechanisms and improving data representation for model discovery.

This paper models the adaptive immune system as a reinforcement learning framework, showing how Th cells learn to optimize immune responses through antigen recognition and regulatory signaling.

This study shows that removing tRNA genes from a chromosome does not broadly disrupt chromosome folding but affects local nucleosome positioning and SMC protein binding, as well as centromere and heterochromatin clustering.