About Me
Hi, my name is Siyuan Jiang. I am an undergraduate student at 
Tsinghua University, Tanwei College, majoring in Chemical Biology for Pharmaceutical Science.
My research journey began with synthetic biology on the 
Tsinghua-M iGEM 2024 team, where we built RNAssay, an ADAR-based RNA sensing system in yeast, and won a Gold Medal. After a brief stay in Prof. Xuebin Liao’s lab, I joined Prof. Yigong Shi’s lab at Tsinghua University in July 2025, where I study the evolution of the gamma-secretase complex. In summer 2026, I visited 
Prof. Le Cong’s lab at Stanford University, and in September 2026 I visited 
Prof. Xuefeng Liu’s lab at the University of Florida.
My research interests center on computational protein engineering, especially model-driven protein design, virtual screening, evolutionary analysis, and functional redesign.
Education
Tsinghua University, Tanwei College
- GPA: 3.83/4.00
- Selected coursework: Chemical Bioinformatics and Artificial Intelligence (A+), Drug Design (A+), Gene Editing (A+), Bioinformatics (A, Excellent Project), Medicinal Chemistry (A), AI in Healthcare (A), Chemical Biology (A-)
Research Interests
My primary research interests include:
- Protein Design Models: Developing generative and predictive models for protein sequence, structure, and function by integrating structural priors, evolutionary information, and experimental feedback.
- Virtual Screening: Building efficient screening pipelines for protein and peptide candidates to reduce computational bottlenecks before high-precision structural validation.
- Functional Engineering: Using evolutionary and functional analyses to guide rational modification of enzymes, binders, and other engineered proteins.
Research Experience
Structure-Guided Aptamer Design
- Developed workflows for RNA/DNA aptamer design, integrating protein–nucleic acid complex prediction, backbone-conditioned inverse folding, and structure-based candidate filtering.
- Benchmarked AlphaFold 3, Boltz-2, Protenix, and OpenFold3, evaluating nucleic acid folding, protein-relative binding poses, and interface quality to assess the reliability of predicted training structures.
- Evaluated NALite inverse-folding models with predicted and generated structural data under matched training budgets, comparing RNA/DNA sequence recovery and data-source effects.
EasyDesign: Agent-Assisted Protein Design
- Contributed across multiple stages of EasyDesign’s agent-assisted protein and binder design workflow.
- Primarily responsible for optimizing structure-prediction models and inference workflows; configured an inference backend using AlphaFold 3 code with converted OpenFold3 weights and packaged reproducible GPU environments for deployment on laboratory servers.
High-Throughput Screening of Functional Food-Derived Peptides
- Developed a multi-stage screening pipeline for bioactive food-derived peptides targeting specific proteins, addressing the computational bottleneck of direct high-precision docking.
- Constructed and trained a lightweight pre-screening model to filter massive peptide libraries and enrich candidates for downstream analysis.
- Integrated AlphaFold 3 for high-precision structural validation of top candidates and finalized the study with wet-lab experimental assays to verify binding affinity.
Spatiotemporal Evolution of the Gamma-Secretase Complex
- Constructed eukaryotic phylogenetic trees for the gamma-secretase complex and traced the PS1 subunit from prokaryotic homologs to clarify its deep evolutionary origin.
- Batch-predicted approximately 1,500 structures with AlphaFold 3 and analyzed the stepwise assembly order of gamma-secretase subunits during evolution.
- Applied statistical methods and deep learning tools to identify mechanistically important residues in gamma-secretase and explore their potential association with Alzheimer’s disease.
Publications
-
IEEE BIBM
-
npj Structural Biology
Selected Projects
Protein-Peptide Interaction Prediction via Generative Docking & GNNs
- Developed a generative-discriminative framework coupling diffusion-based docking (RAPiDock) with a Transformer-GNN scorer (ITN).
- Implemented multi-instance learning on 3D bipartite graphs to enable structure-aware, interpretable binding prediction.
- Outperformed sequence-based baselines in AUC and enrichment on pMHC I and SH3-peptide systems.
Honors and Awards
- WeiGuang Program, Individual Excellence Award, Tsinghua University (Jun 2025)
- Gold Award for the Practical Detachment, Tanwei College (Jan 2025)
- Comprehensive Excellence Award, Tsinghua University Scholarship, Top 20% (Nov 2024)
- Gold Medal, International Genetically Engineered Machine Competition (iGEM), team Tsinghua-M · RNAssay (Oct 2024)
Skills
- Programming: Python, C++, R, Java, LaTeX
- Frameworks: PyTorch
- Languages: English (TOEFL 94)
Hobbies
Beyond research, I maintain an active lifestyle and diverse interests:
- Sports: I am a member of the Tanwei College Badminton Team and the Tsinghua University Diving Association. I also enjoy running and swimming.
- Arts & Culture: I love Rock Music, J-Pop, and Sci-Fi films (e.g., The Matrix, Star Wars).
Recent Photos
)
