Dr. Huang's research focuses on developing advanced computational architecture (machine learning and deep learning) and statistical methods to integrate multiple types of omics data, including DNA-seq(Whole genome sequencing), RNA-seq, Chip-seq, Methylation, large-scale pharmaco-genomics data and in-house electronic health records, for driver signaling pathway analysis, drug repositioning, drug combination prediction, stromal/immune-tumor cell communication and drug-resistant mechanisms and validating the computational discoveries using in-vivo xenograft models. Meanwhile, He has been collaborating closely with clinicians from different disciplines (Cancer, Immunology, developmental, and Neurology) to seek the computational solutions to the cutting-edge medical problem arising from clinical practice.
Dr. Huang has designed a non-parametric, bootstrapping based simulated annealing (NBPSA) approach to identify driver signaling pathways or predict cell-cell interactions for cancer patients by integrating multiple personal genomic profiling (DNA-seq, RNA-seq, Copy number, Chip-seq, Methylation). This work sheds light on driver signaling pathways with genes that show simultaneous mutation, methylation, copy number, and gene expression alterations are likely to play important roles in tumor metastasis and drug-resistance. Based on the driver signaling pathway analysis, He has developed a DSNI-DFN approach for drug repositioning. DSNI-DFN identified Cardiac Glycosides that inhibit the Groups 3 and 4 medulloblastoma tumor from 1,309 candidate drugs. One of the Cardiac Glycosides, digoxin, is validated in Groups 3 and 4 medulloblastomas in vivo. He has also developed two systems biology platforms, i.e. DrugComboRanker and DrugComboExplorer, in order to reduce drug resistance in cancer therapy and identify the mechanisms of drug synergism. These drug combination tools have demonstrated a strong predictive ability for specific cancer types (non-small cell lung cancer, estrogen receptor-positive breast cancer, Ewing Sarcoma, and AR-negative prostate cancer). The predicted drug combinations are further validated by in-vivo experiments. Meanwhile, such work uncovers the potential mechanisms of action of drugs and synergistic effects of drug combinations from a driver signaling pathway perspective.