Zhenzhi Tan

Hi, I am Zhenzhi Tan, a Ph.D. student in organic chemistry at the Center of Basic Molecular Science, Department of Chemistry, Tsinghua University, advised by Prof. Sanzhong Luo. My research focuses on AI4Science, especially data-driven organic reaction discovery, active learning optimization and reliability for these models.

CN · Chinese CV · English CV

Education

• CBMS, Department of Chemistry, Tsinghua University, Ph.D. student in Organic Chemistry (AI4Science), Advisor: Prof. Sanzhong Luo, Sept. 2024 - Present
• Department of Chemistry, Tsinghua University, B.S. in Chemistry, GPA: 3.90/4.00 (Rank: 2/25), Graduate of Tsinghua University, Sept. 2020 - July 2024

Research Experience

Data-Driven Discovery & Active Learning for Orangic Synthesis
Project leader (NSFC Ph.D. project), Mar. 2024 - Present

• Developed Bayesian optimization algorithms tailored for small-sample, high-noise organic synthesis.
• Implemented an automated DFT descriptor extraction pipeline integrated with Bayesian optimization.
• Achieved global optimization of a 1.8-million reaction space within only 63 experiments, significantly accelerating reaction discovery.

SynAD Reliability Framework for Ullmann Reaction Prediction
Project leader (NSFC undergraduate project), Sept. 2023 - Oct. 2025

• Built a high-accuracy Ullmann reaction yield prediction model by integrating physical organic chemistry knowledge with automated DFT-based descriptor computation.
• Proposed the Synthetic Applicability Domain (SynAD) framework to address the extrapolation trap of AI in chemical synthesis and establish uncertainty assessment in unknown chemical spaces.
• Validated the framework on 7 in-house and public datasets, showing that SynAD effectively filters unreliable predictions and improves model reliability.

Computer Vision for Polymer Miscibility Prediction
Second author

• Constructed a deep convolutional neural network (CNN) visual model to automatically extract high-level features from SEM microscopic images of polymer blends.
• Enabled rapid automatic characterization of miscibility and related material properties, replacing inefficient manual analysis workflows.

Skills

• Programming: Python (PyTorch, Scikit-learn, Numpy, Pandas), C++ (basic), SQL, Linux/Bash, Git
• AI4Chem: Cheminformatics with RDKit and DeepChem; DFT calculations with Gaussian, ORCA, and xTB

Honors & Awards

• Participant, 74th Lindau Nobel Laureate Meeting, 2025
• Best Poster Award, Nature Conference on Automation for Chemistry, 2025
• Best Poster Award, International Conference of Computational Organic-synthesis Catalysis, 2025
• T. P. Hou Award, 2024
• Outstanding Graduate of Tsinghua University, 2024
• National Scholarship, 2022
• Tsinghua University Comprehensive Scholarship, 2021
• Gold Medal, 33rd Chinese Chemistry Olympiad (CChO), 2019