Our goal is to leverage multi-omics technologies in conjunction with bioinformatics, AI, and systems biology approaches to decipher the fundamental principles and regulatory mechanisms governing cell identity formation, maintenance, and transition across diverse biological processes in both physiological and pathological states.
We pioneer the development of novel multi-omics technologies and computational tools to enhance our capacity for unraveling the genetic and epigenetic mechanisms underlying development, immunity, and cancer. Through close collaboration with clinicians, we apply our omics approaches to address basic medical and clinical questions in fields such as dermatology, stomatology, and oncology.
Below are highlights of our ongoing projects:
Human skin cell atlas
We employ cell atlas methodologies to investigate human skin cell diversity across anatomical sites, aging processes, and various healthy/diseased states—including skin cancers. Our research includes bulk and single-cell transcriptomic profiling:
Integration with multi-omics profiling of basal cell carcinoma (BCC) and extramammary Paget’s disease (EMPD) revealed a universal cancer predisposition principle: Transcription factors with high baseline expression in site-specific cells-of-origin (e.g., SOX11 in facial hair follicle cells for BCC; PITX1 in penile keratinocytes for EMPD) drive malignancy via feedback-amplification circuits that lower oncogenic thresholds.
Using multi-omic approaches (epigenome, transcriptome, proteome) and single-cell resolution analysis, we mapped the developmental trajectory from precancerous goblet cells to hillock cells and finally Paget cells. We identified FOXA1 as a driver of hillock cell fate, validated in HaCaT cells and transgenic mouse models, with FOXA1-mediated chromatin accessibility and super-enhancer epigenetic reprogramming as the underlying mechanism.
Our single-cell RNA-seq analysis of 218,021 cells from 20 plantar melanoma and 6 normal samples—combined with spatial transcriptomics, mIHC, single-cell multiomics, and bulk RNA-seq survival analysis—revealed that plantar melanoma evolves from stem cells through transitional and Schwann cell-like precursor states to a terminal Schwann cell-like state regulated by HMGA2. This state is associated with immune cell dysregulation, poor prognosis, and enhanced invasion/lymph node metastasis; inhibiting HMGA2 blocks this transition (Tian et al., Theranostics, 2025).
New single-cell epigenomic technologies
We are developing innovative single-cell multi-modality sequencing technology, enabling simultaneous measurement of gene expression (RNA), chromatin accessibility (ATAC), and histone modifications (CUT&Tag) in individual cells. The technology is scalable to process up to 1 million cells across multiple samples, covering major histone marks (H3K4me1/3, H3K27ac, H3K27me3, H3K9me3) in a single experiment. Optimized for clinical samples (e.g., core needle biopsies), it provides high-resolution epigenomic profiling for investigating epigenetic regulation in cancer.
New bioinformatics tools and databases. We develop advanced bioinformatics methods for genome and sequencing data analysis, including:
We are also developing specialized methods for analyzing our newly developed single-cell epigenomic multi-modality sequencing datasets.
Pan-cancer epigenomic profiling
Using our novel single-cell multi-omics approaches, we are profiling epigenomic landscapes across diverse cancers (breast, liver, gastric, colorectal, etc.). Our goal is to characterize common and cancer-type-specific epigenomic features, ultimately deciphering the underlying epigenetic mechanisms of cancer development.