EXPERTISE
Genomic epidemiology, Risk prediction model, Bayesian statistics, Statistical consulting
RESEARCH INTERESTS
High-throughput omics data analysis (including single-cell RNA sequencing)
– Identifying and removing batch effects
– Identifying biomarkers (Early detection of disease, diagnosis, and prognosis)
– Pathway-level biological interpretation
Immunoncology or tumor microenvironment
– Immune cell infiltration signatures
– Tumor purity, stromal signature and immune signature
Statistical consulting
– Consulting services to conduct study design and to present statistical results
– Publication-ready figures
RESEARCH ACTIVITIES & ACCOMPLISHMENT
Dr. Chen is a biostatistician who provides consulting services including experimental design, and data analysis and interpretation. In collaboration with the principal investigators, Dr. Chen has analyzed several public datasets to provide further validation, and has organized a related manuscript for publication. His previous researches are described below.
1. Reproducibility of high-throughput data. Genomic epidemiology provides the opportunity to identify biomarkers for precision medicine, but the reproducibility of these studies still presents challenges, and relies on making extensive and intensive use of principal component analysis. Examples include dealing with population substructure to avoid spurious genetic associations in genome-wide association studies (GWAS), measured and unmeasured batch effects in high-throughput data, such as NanoString panels (gene expression arrays), DNA methylation arrays, and single-cell RNA sequencing. Another strategy to improve the reproducibility of these studies is gene set enrichment analysis (GSEA). With the aid of GSEA, many studies have successfully achieved disease prevention and intervention strategies.
2. Risk prediction model. Lung cancer in never-smokers is the 7th most common cause of cancer-related death worldwide. Identifying a high-risk group in this population is the key to cancer prevention. GWAS identified common genetic variants that contribute to lung cancer susceptibility in Asian never-smoking females (NSF). Thereafter, these genetic and environmental risk factors were used to construct a prediction model, the Taiwan NSF lung cancer risk models using genetic information and simplified questionnaire (TNSF-SQ). The TNSF-SQ could be used to predict lung cancer occurrence in Taiwan NSF and identify high-risk Taiwan NSF who may benefit from low-dose CT screening.
3. Training workshops. Dr. Chen is one of the speakers who teach several topics on analyzing high-throughput data in bioinformatics training workshops. In 2017 and 2018, more than 330 participants attended the workshops in which he was involved.
