Wang Lab

Digital Intelligence for Tissue Regeneration, ZOC and GIBH

Research Interests

     Evolutionary selection has shaped striking diversity in regenerative capacity across species and tissues. The Wang Lab focuses on elucidating the principles underlying these differences in regenerative capacity, particularly in the retina and liver. While zebrafish and other non-mammalian species can regenerate the retina, this capacity is largely absent in mammals. In contrast, mammals have retained a remarkable ability for liver regeneration, yet chronic injury can impair this regenerative program and lead to fibrosis. How have these differences in regenerative capacity evolved? Can regenerative programs from regenerative species be harnessed to restore retinal regeneration in humans? Can liver regeneration be enhanced to prevent or reverse fibrosis?
     To address these questions, we integrate high-throughput sequencing and organoid technologies with statistical and computational approaches, complemented by cellular and animal experiments. Through these interdisciplinary approaches, the Wang Lab is dedicated to uncovering fundamental principles of tissue regeneration and developing regenerative strategies for diseases such as tissue fibrosis and degeneration.

Biological Interests

Penatibus

Tissue Regeneration

     Endogenous tissue regeneration harnesses the intrinsic regenerative potential of resident cells to restore tissue architecture and function after injury. In mammals, regenerative capacity varies greatly across tissues. For example, the mammalian retina has very limited regenerative capacity, whereas the liver can regenerate efficiently following acute injury.

     Retinal regeneration is diverse across species. In zebrafish, Müller glia (MG) can be reprogrammed to generate all retinal cell types after injury. In contrast, mammalian MG do not induce an effective regenerative response. By comparing zebrafish and mouse retinas using single-cell RNA sequencing and functional experiments, we identified core regulatory networks and key regulators, including Nfia/b/x (Hoang et al., Science, 2020; Jiang et al., iScience, 2022). Notably, MG-specific knockout of Nfia/b/x promotes the generation of amacrine and bipolar cells in mouse retinas.

     The liver, in contrast, can regenerate efficiently following acute injuries, including partial hepatectomy and drug-induced acute injury. Liver regeneration depends primarily on hepatocyte proliferation and functional recovery. Multiple signaling regulators, including YAP, HBEGF, and HGF, promote hepatocyte proliferation during regeneration. However, how epigenetic regulation and interactions among non-parenchymal liver cells control liver regeneration remain poorly understood.

Tissue Fibrosis

     Tissue fibrosis is characterized by excessive accumulation of extracellular matrix (ECM) during tissue repair following chronic injury. It affects multiple tissues, including the liver. Although the liver can undergo scar-free repair and regeneration following acute injury, chronic injury leads to persistent ECM deposition and progressive fibrosis. Most liver cell types contribute to tissue repair, with hepatic stellate cells (HSCs) serving as the main source of ECM. However, how different cell types, particularly HSCs, regulate transitions between regenerative repair and fibrosis remains poorly understood. Understanding these mechanisms will facilitate the development of strategies to prevent or reverse liver fibrosis.

     To investigate these mechanisms, we integrated high-throughput sequencing datasets to characterize HSC heterogeneity in mouse and human livers. We identified eleven HSC subtypes, including quiescent, activated, and inactivated HSCs. We found that LHX2, which is highly expressed in quiescent and inactivated HSCs, promotes hepatocyte proliferation by regulating HGF transcription (Tao et al., Hepatology, 2025). Moreover, LHX2 suppresses liver fibrosis by modulating SMAD6 expression.

Computational Interests

Elementum Tempus

Omics and Regulatory Networks

     Trans-acting factors, cis-regulatory elements, and epigenetic modifications regulate the expression of target genes. These components and their regulatory relationships form gene regulatory networks (GRNs). GRNs are valuable to elucidate regulatory mechanisms underlying diverse biological processes, including tissue regeneration. With the development of single-cell sequencing and spatial transcriptomics, GRNs can be inferred at cellular and spatial resolution. However, it remains challenging to accurately infer GRNs from these omics data and elucidate the underlying biological mechanisms.

     We therefore developed several algorithms for GRN inference and analysis, including IReNA, CACIMAR, and ScReNI (Jiang et al., iScience, 2022; Jiang et al., Briefings in Bioinformatics, 2024; Xu et al., Genomics, Proteomics & Bioinformatics, 2025). Through integrating single-cell RNA sequencing (scRNA-seq) and single-cell assay for transposase-accessible chromatin using sequencing (scATAC-seq) data, IReNA and ScReNI reconstruct modularized and cell-specific GRNs, respectively. CACIMAR enables cross-species comparison of GRNs and identifies evolutionarily conserved regulatory modules. We have applied these approaches to reveal key regulators and regulatory mechanisms underlying tissue regeneration and development (Hoang et al., Science, 2020; Xu et al., Genomics, Proteomics & Bioinformatics, 2025).

Omics and Artificial Intelligence


     Artificial intelligence, particularly deep learning, provides powerful approaches for uncovering biomedical mechanisms by learning complex relationships from high-throughput omics and other multimodal data. With the rapid development of scRNA-seq and spatial transcriptomics technologies, large-scale and increasingly diverse omics data are becoming available. Conventional statistical and machine learning methods have been widely applied to analyze these data and generate valuable biological insights. However, many biological processes involve complex and nonlinear relationships across multiple data modalities, including omics and medical imaging data. Therefore, new computational approaches, particularly deep learning methods, are needed to effectively integrate multimodal data and uncover complex regulatory mechanisms underlying biomedical processes, including tissue regeneration.