Ruizhi Deng

The paper develops MethaDory, a framework to train and validate DNAm-based classifiers for rare diseases using synthetic data, achieving high accuracy across multiple platforms.

This study identifies a human-specific CHD2 regulatory variant that influences cortical development tempo via lysosomal homeostasis, showing therapeutic potential through lysosomal restoration.

The paper develops BRAIN-MAGNET, a convolutional neural network that predicts non-coding regulatory element activity and helps prioritize disease-causing variants in neurogenetic disorders.

The paper identifies 39,709 differentially active enhancers in human fetal brain development linked to neurodevelopmental disorders and GWAS loci.