01

About Me

I am an undergraduate in bioinformatics at the School of Life Sciences, Peking University, pursuing a double degree in economics at the National School of Development.

In summer 2026, I conducted research in the Qiu Lab at Stanford, where I worked on Bio-Babel and developed agentic and end-to-end approaches for imaging-based spatial transcriptomics.

Outside research, I enjoy playing volleyball.

02

Research Interests

  • Spatial Multi-omics spatial transcriptomics and translatomics, multimodal analysis, and spatial oncology
  • Scientific Agents agentic workflows for complex analyses and biological discovery
  • AI4Science deep learning for more accurate methods and virtual cell and embryo modeling
  • Bioinformatics Software reliable tools, interoperable packages, and reproducible workflows

03

Selected Work

GitHub profile
  1. In progress

    Spatial Multi-omics Analysis of GBM with STATES

    STATES simultaneously profiles the spatial transcriptome and translatome, enabling a joint view of RNA abundance and active translation in tissue. We are optimizing the upstream processing pipeline and conducting downstream multimodal analyses to investigate the spatial organization and translational regulation of GBM.

    Spatial multi-omics · Spatial oncology · GBM

  2. Manuscript in preparation

    Bio-Babel

    Bio-Babel is an agentic framework that enables AI agents to understand, use, and translate scientific software. We contributed across Agent Build and Agent Read through automated package annotation, Codex integration, workflow validation, and cross-language extensions. Co-author of the accompanying manuscript.

    Scientific agents · Software translation · Research software

  3. Stanford Summer Research

    CycLENS — End-to-End iST Model

    We are developing CycLENS, a unified model that jointly learns image alignment, transcript localization, and barcode decoding from raw multiplexed fluorescence images. By replacing fragmented upstream processing with joint learning, we aim to reduce error propagation and generalize across experiments, imaging platforms, and iST technologies.

    Deep learning · iST · Multimodal learning · Bioimage analysis

  4. In development

    iST Upstream Processing Agent

    We are developing a scientific agent for iST upstream processing that automates data organization, quality control, parameter tuning, algorithm selection, and workflow orchestration. It is designed to help researchers configure and optimize complex pipelines across technologies and platforms, including registration, data augmentation, spot detection, decoding, and cell segmentation.

    Scientific agents · iST · Bioimage processing · Workflow orchestration

04

Contact

I am always happy to connect—whether you would like to discuss research, share an idea, collaborate, or simply say hello.