Francesca-Zhoufan LiSearch Francesca-Zhoufan Li Francesca-Zhoufan Li Home Experience Publications Events Contact Light Dark Automatic Francesca-Zhoufan Li AI for Science & Engineering, currently focusing on machine learning for proteins California Institute of Technology Arnold Lab Yue Group Amazon AI4Science Fellow NSF Graduate Research Fellowship Biotech Leadership Training Program About With a broad interest in applying AI to science and engineering problems, I am currently focusing on the development and evaluation of machine learning-assisted protein engineering tools as a Bioengineering Ph.D. student at Caltech, co-advised by Frances Arnold and Yisong Yue. My current main project involves developing zero-shot predictors for non-native enzyme activity prediction, building on my recent work, Evaluation of Machine Learning-Assisted Directed Evolution Across Diverse Combinatorial Landscapes. I have also worked with Kevin K. Yang, Alex X. Lu, and Ava P. Amini through my summer internship at Microsoft Research New England on Feature Reuse and Scaling: Understanding Transfer Learning with Protein Language Models, which was presented at ICML 2024. I am passionate about making computational tools usable and accessible to the wet-lab scientists and engineers. In collaboration, I contributed to developing data analysis and interactive visualization software for Long-read every variant Sequencing (LevSeq), a newly developed nanopore-based technology for rapid protein sequencing. During and shortly after my time at the University of California, Berkeley where I earned my B.S. in Bioengineering and my B.S. in Chemical Biology, I gained experience in various research areas: developing RNA-seq software tools at Zymergen, discovering genetic circuit components with Richard Murray, contributing to cancer immunotherapy and SARS-CoV-2 antibody therapeutics development with Shohei Koide,optimizing cell-free platforms at Tierra Biosciences, and undertaking metabolic engineering and synthetic biology tool building projects at the Dueber Lab. Outside of research, I enjoy being active outdoors, experiencing diverse cultures, solving fun puzzles, and doing minimalism iPhoneography. Given my personal background and journey, I am committed to promoting equitable opportunities and individualized education in STEM through mentoring, teaching, and outreach volunteering. Download my resumé . Interests Machine learning Protein engineering Biomolecular tool development Bioengineering Interactive visulization Education Ph.D. in Bioengineering, 2025 California Institute of Technology B.S. in Bioengineering, 2019 University of California, Berkeley B.S. in Chemical Biology, 2019 University of California, Berkeley Experience Highlights BioML Research Intern Microsoft Research Jun 2022 – Sep 2022 Cambridge, MA Transfer learning for pretrained protein language models Machine Learning for Proteins Arnold Lab & Yue Group, Caltech Jan 2021 – Present Pasadena, CA Evaluated 6 general, 10 substrate-aware, and ensemble zero-shot predictors—including generative modeling, molecular docking, and active-site heuristics—across 22 substrates for non-native enzyme activities, leveraging evolutionary patterns, pretrained language and structural models, inverse folding models, generative models, molecular docking, stabilities, and other heuristics Systematically analyzed multiple machine learning-assisted directed evolution strategies, including active learning and focused training using six distinct zero-shot predictors, across 16 diverse protein fitness landscapes Facilitated the rapid generation of sequence-function data and tool development for constructing protein mutant libraries Featured Publications Substrate-Aware Zero-Shot Predictors for Non-Native Enzyme Activities Substrate-aware zero-shot predictors can accelerate enzyme engineering by estimating activity for non-native substrates and new-to-nature reactions, with a weighted ensemble of AlphaFold 3 and EVmutation scores demonstrating broad generalization across diverse chemistries. Francesca-Zhoufan Li, Lukas A. Radtke, Kadina E. Johnston, Cheng-Hao Liu, Yisong Yue, Frances H. Arnold PDF Cite Code LevSeq: Rapid Generation of Sequence-Function Data for Directed Evolution and Machine Learning LevSeq is a nanopore sequencing-based pipeline that rapidly generates sequence-function data for entire protein-coding genes, integrating into protein engineering workflows to enable data-driven directed evolution and machine learning-guided protein engineering. Yueming Long, Ariane Mora, Francesca-Zhoufan Li, Emre Gürsoy, Kadina E. Johnston, Frances H. Arnold PDF Cite Code Project DOI Evaluation of Machine Learning-Assisted Directed Evolution Across Diverse Combinatorial Landscapes A systematic analysis of multiple machine learning-assisted directed evolution strategies, including active learning and focused training using six distinct zero-shot predictors, across 16 diverse protein fitness landscapes. Francesca-Zhoufan Li, Jason Yang, Kadina E. Johnston, Emre Gürsoy, Yisong Yue, Frances H. Arnold PDF Cite Code Dataset DOI Feature Reuse and Scaling: Understanding Transfer Learning with Protein Language Models A systematic analysis of protein language model transfer learning via 370 experiments across downstream tasks, architectures, model sizes, model depths, and pretraining time. Francesca-Zhoufan Li, Ava P. Amini, Yisong Yue, Kevin K. Yang, Alex X. Lu PDF Cite Code Dataset DOI Opportunities and Challenges for Machine Learning-Assisted Enzyme Engineering Machine learning can complement enzyme engineering by helping to discover enzymes with desired function and to accelerate the optimization of enzyme fitness. Jason Yang, Francesca-Zhoufan Li, Frances H Arnold PDF Cite DOI Recent & Upcoming Talks Substrate-Aware Zero-Shot Predictors for Non-Native Enzyme Activities Substrate-aware zero-shot predictors can accelerate enzyme engineering by estimating activity for non-native substrates and new-to-nature reactions, with a weighted ensemble of AlphaFold 3 and EVmutation scores demonstrating broad generalization across diverse chemistries. Apr 27, 2025 4:20 PM — 5:20 PM Singapore EXPO Francesca-Zhoufan Li, Lukas A. Radtke, Kadina E. Johnston, Cheng-Hao Liu, Yisong Yue, Frances H. Arnold PDF Code Generative and Experimental Perspectives for Biomolecular Design Co-organize The Generative and Experimental Perspectives for Biomolecular Design (GEM) workshop at ICLR 2025 to foster collaboration between machine learning experts and experimental scientists. Apr 27, 2025 8:50 AM — 5:00 PM Singapore EXPO Cheng-Hao Liu, Jarrid Rector-Brooks, Soojung Yang, Sidney L Lisanza, Francesca-Zhoufan Li, Hannes Stark, Jacob Gershon, Lauren Hong, Pranam Chatterjee, Tommi Jaakkola, Regina Barzilay, David Baker, Frances H. Arnold, Yoshua Bengio Feature Reuse and Scaling: Understanding Transfer Learning with Protein Language Models We conducted the most comprehensive transfer learning study for protein language models to date and found that scaling pretraining does NOT scale downstream task performance, except for structure prediction Jul 25, 2024 11:30 AM — 1:00 PM Messe Wien Exhibition Congress Center Francesca-Zhoufan Li, Ava P. Amini, Yisong Yue, Kevin K. Yang, Alex X. Lu PDF Code Slides Video Generative and Experimental Perspectives for Biomolecular Design Co-organize The Generative and Experimental Perspectives for Biomolecular Design (GEM) workshop at ICLR 2024 to foster collaboration between machine learning experts and experimental scientists. May 11, 2024 8:50 AM — 5:00 PM Vienna, Austria Chenghao Liu, Jarrid Rector-Brooks, Jason Yim, Soojung Yang, Sidney Lisanza, Francesca-Zhoufan Li, Pranam Chatterjee, Tommi Jaakkola, Regina Barzilay, David Baker, Frances Arnold, Yoshua Bengio Leveraging Sequence and Structure for Machine Learning-Assisted Protein Engineering A poster presentation on multi-modal representation learning for predicting top protein fitness from combinatorial and random mutagenesis libraries May 23, 2023 10:30 AM — May 25, 2023 4:00 PM Oakland Marriott Francesca-Zhoufan Li, Jason Yang, Kadina Johnston, Yisong Yue, Frances Arnold See all events Contact [email protected] Google Scholar Github Connect on LinkedIn DM on Bluesky DM on Twitter © 2025 Francesca-Zhoufan Li. Published with Wowchemy — the free, open source website builder that empowers creators. Cite Copy Download