Representative Projects From computational inference to experimental validation and engineering implementation

This project systematically dissects cell-cell communication structures across cancer types to identify conserved regulatory circuits, cancer-specific signaling axes, and communication hub populations with important topological roles.

• Built standardized single-cell RNA-seq analysis pipelines covering more than 10 cancer types to characterize tumor microenvironment composition and communication networks.
• Combined CellChat and NicheNet to identify conserved and cancer-specific ligand-receptor regulatory circuits.
• Reimplemented CoVarnet in Python and identified five modules associated with cancer development.
• Integrated spatial transcriptomics with scRNA-seq to localize high-confidence signaling niches and validate spatial co-localization among sender cells, receiver cells, and pathway activity.
• Identified communication hub populations using centrality and network-topology metrics, and linked hub-driven pathways to tumor-associated phenotypic programs.

This project uses a doxycycline-regulated KRAS-mutant pancreatic cancer model to investigate tumor recurrence after KRAS inactivation and the molecular transition toward a RAS-independent disease state.

• Investigated a doxycycline-inducible KRAS-mutant pancreatic cancer model and observed tumor recurrence after doxycycline withdrawal and KRAS inactivation.
• Analyzed RNA-seq data from recurrent tumors and identified upregulation of EMT-related transcriptional programs.
• Prioritized EMT-associated gene programs as a candidate mechanism of RAS-independent recurrence and as a basis for subsequent functional validation and therapeutic target discovery.

This project combines function-specific mutant construction with the dTAG rapid-degradation system to determine whether TP53RK function in medulloblastoma primarily depends on phosphorylation-related activity or KEOPS-complex-associated mechanisms.

• Constructed plasmids carrying TP53RK phosphorylation-site mutants and KEOPS critical-site mutants.
• Introduced mutant constructs through lentiviral transduction and used dTAG-V1 to degrade endogenous TP53RK.
• Used Incucyte live-cell imaging to quantify proliferation and compare rescue differences in medulloblastoma cell-death phenotypes between the two mutant classes.

TasAnchor is a modular whole-cell adhesion and immobilization platform built in Bacillus subtilis, integrating genetic engineering, protein-scaffold design, and bioreactor membrane immobilization.

• Served as the current team leader of SCU-China; the 2025 project received an iGEM Gold Medal and was named an Undergraduate Top 10 Project.
• Knocked out endogenous tasA in Bacillus subtilis and constructed a pHT01-based SpyTag/SpyCatcher modular adhesion scaffold.
• Expressed and purified recombinant proteins in E. coli BL21(DE3) to validate the modular immobilization system.
• Immobilized engineered Bacillus subtilis biofilms on polystyrene bioreactor membranes to demonstrate whole-cell immobilization performance.