Shunichiro Tomura

This paper improves genomic prediction accuracy for maize flowering-time traits using an ensemble approach, revealing novel insights into trait genetic variation.

This study evaluates graph attention networks (GAT) for genomic prediction in crops, finding that ensembles of GAT models outperform single models using gene network structures.

This study evaluates three weight optimisation approaches for genomic prediction ensembles in maize, finding that weighted models improve prediction performance when weights differ from equal weights, but no single method clearly outperforms others.

EasiGP is a tool that interprets genomic prediction models by visualizing marker effects on trait phenotypes, revealing both known and novel genetic regions in maize.

The paper shows that combining multiple genomic prediction models improves accuracy and reduces errors in predicting crop traits, supporting the Diversity Prediction Theorem.