Haoyang Zeng

The paper uses deep learning on images of iPSC-derived motor neurons with fALS mutations to identify disease-relevant phenotypes and molecular differences.

This paper introduces a novel KSM motif representation and KMAC method for de novo motif discovery, showing improved accuracy in predicting TF binding sites and regulatory variant effects.

CpGenie predicts allele-specific DNA methylation and identifies methylation quantitative trait loci using deep learning.

EnsembleExpr is an ensemble-based framework that accurately predicts eQTLs using MPRA data and outperforms existing methods in a CAGI challenge.

GERV predicts regulatory variants affecting transcription factor binding by learning a k-mer based generative model from ChIP-seq and DNase-seq data, outperforming existing methods in SNP prediction.