Xiuru Dai

The paper develops a sequence-based classifier to identify phenotype-associated genes in plants, validated across maize, rice, and arabidopsis.

This study uses ChIP-seq to map TF binding in maize leaves, reconstructs a regulatory network covering 77% of expressed genes, and shows its conserved evolution across species.

This paper evaluates machine learning algorithms for predicting gene functions without relying on homology, finding random forest provides the most accurate results, particularly in Biological Process GO terms.