About & CV
My doctoral research focuses on developing machine-learning methods and research tools for therapeutic peptides, bringing together literature evidence, traceable data and molecular representations. My work combines computational research with full-stack software development.
East China University of Science and Technology
Integrated master’s–doctoral programme, School of Pharmacy. Supervisor: Professor Yun Tang.
Coursework rank: 1st of 255 students in the same entering cohort of the School of Pharmacy; not an overall research assessment.
Beijing University of Chemical Technology
Undergraduate studies in pharmaceutical engineering.
Three-year undergraduate comprehensive assessment: ranked 3rd of 80, combining coursework, research and extracurricular activities.
Beyond model development, I build the interfaces, APIs and computing workflows that make research accessible. I have three years of experience administering laboratory computing infrastructure and managing more than 20 research websites and services.
Managed one control node and 12 compute nodes, including four GPU nodes with 11 GPUs. Supported admetSAR platforms whose lifetime visit counts exceed one million across versions 1.0–3.0, including over 170,000 for version 3.0 as of October 2026. These are platform cumulative counts, not unique users or traffic generated personally.
admetSAR 3.0 ↗I connect computational research across small molecules, peptides and proteins with data workflows, full-stack software development and the infrastructure needed to run them.
Data cleaning, feature construction and pretrained representations; predictive model development, training, evaluation and visualisation. In Bilingua, I implemented and evaluated the fusion of peptide sequence and molecular representations.
Methods & toolsPython · PyTorch · ESM-C · Uni-Mol2
Connect backbone generation and sequence design with structure prediction and candidate evaluation. My antibody work includes CDR redesign, multiple-seed sampling, interface analysis and comparison of modelling settings to support experimental prioritisation.
Methods & toolsRFantibody / RFdiffusion · ProteinMPNN · AlphaFold2/3 · RoseTTAFold2
Experience spans small-molecule pocket analysis and screening, peptide activity/property prediction, and protein/antibody structure analysis. I have used docking and empirical affinity estimates, contributed to Rosetta pocket design, and participated in a team screening workflow using Glide and MM/GBSA.
Methods & toolsPyMOL · Rosetta · Schrödinger · HADDOCK · PRODIGY · Glide / MM/GBSA
Integrate LLM APIs and prompts into structured extraction workflows with source-linked evidence and layered evaluation. PepMinerva covers retrieval, full-text parsing, routing, extraction and export, with a manually annotated benchmark, CLI and Agent Skill interface.
Methods & toolsLLM APIs · Structured extraction · Benchmarking · CLI · Agent Skill
Turn data processing and model inference into backend services, APIs and usable interfaces. I develop and deploy research websites, webservers and desktop tools; PepDiana connects source-aware data structures with browser and programmatic access.
Methods & toolsData services · APIs · Web interfaces · Desktop applications
Three years of Linux and CPU/GPU cluster administration, Slurm batch workflows, web-server configuration and service maintenance. Managed more than 20 research sites/services and maintained practical guidance for laboratory users.
Methods & toolsLinux · Slurm · CPU/GPU clusters · Nginx / Apache
Selected university and postgraduate honours; team awards are identified explicitly.
Chinese: native. English: CET-6 passed.
Get in touch
I am interested in research in computational design of molecules, therapeutic peptides or proteins. I welcome enquiries about research, collaboration or employment opportunities.
2145573258@qq.com ↗