Yushiro Fuji

The paper uses interpretable machine learning to analyze multi-omics data from soybean under drought stress, identifying key biomarkers and interactions influencing plant phenotypes.

The paper proposes a nonlinear two-step model integrating proxy intermediate omics traits to improve phenotypic prediction in soybean, showing better accuracy than linear models.

I-SVVS improves clustering accuracy and reduces computation time for multi-omics microbiome data across soybean, mice, and human datasets.