Welcome!
Email: jjian [at] scripps [dot] edu
Hi there! I’m Jinglin Jian (简靖琳), a PhD student at Scripps Research by the beautiful ocean 🏖️ at San Diego, CA. I’m deeply grateful to be supported by the Kellogg Fellowship, a three-year endowed award generously funded by the Kellogg family and The ALSAM Foundation.
I received my master’s degree from the School of Information Sciences at the University of Illinois at Urbana-Champaign, where I had the opportunity to work closely with Professor Qingyun Wang, Professor Haohan Wang, and Professor Ge Liu. Previously, I studied B.S. in Computer Science and Education at Beijing Normal University, with a dual B.Econ. degree in Economics from Peking University. I have a general interest in how people learn[1] and how computers transform human thinking[2].
I’m also the founder of the AI × Science Club, a student-led local community exploring how AI is reshaping scientific discovery. We’re still growing and building things out, so if this sounds interesting to you, come join us on Slack — we’d love to have you!
News
- Feb 2026: Founded the AI × Science Club at Scripps Research. We’re still building, if you’re interested, come join us on Slack!
- Feb 2026: Rotating in Bryan Briney’s lab, working on antibody language models.
- Dec 2025: Tutorial on “Human-AI Co-Discovery” submitted to ACL 2026.
- Dec 2025: Survey paper “Exploring Agentic Multimodal LLMs for AI Scientists” published on TechRxiv.
- Oct 2025: Selected for the three-year Kellogg Fellowship at Scripps Research. See here.
- Aug 2025: Started PhD study at Scripps Research, rotating in Stefano Forli’s lab.
- May 2025: Graduated from UIUC with MS in Information Sciences.
- Nov 2024: Two papers accepted to IEEE BigData 2024, including one as first author.
Research Interests
I am broadly interested in AI for Science.
RI1. Generative Models for BioMedicine. Developing generative models for biomedical applications. Currently learning diffusion models and flow matching, and working with different architectures on antibody sequence data.
RI2. Agentic Multimodal Large Language Models for Scientific Discovery. Recently working on Agentic AI Scientist for different biomedical discovery processes.
Publications and Conferences
The AI Scientist in Health: Potential, Challenges, and the Road Ahead
Advised by Prof. Qingyun Wang (William & Mary) and Prof. Qingyu Chen (Yale)
A paper discussing the potential, challenges, and future direction of AI Scientists in health systems.
Allosterically Inhibiting Pseudomonas aeruginosa's Polyphosphate Kinase 2A by Disrupting Its Oligomerization
Constanza Torres-Paris, Madeline G. Ammend, Joseph Agha, Jinglin Jian, Matthew Holcomb, Stefano Forli, Lisa R. Racki
Performed virtual screening of 2.4 million compounds against Ppk2A mutant proteins using AutoDock-GPU and identified preliminary hit candidates.
Tutorial: Human-AI Co-Discovery
Qingyun Wang, Wenpeng Yin, Lifu Huang, Yi R. (May) Fung, Jinglin Jian, Xuehang Guo, Ruochen Li
This tutorial is currently under submission to ACL 2026.
Patient Outcome Predictions via A Multimodal Language Model for Electronic Health Records
Zihan Li, Jinglin Jian, Chundian Li, Jinxia Yao, Jin Chen, Yang Zhang
Early prediction of mortality risk and hospital length of stay is critical. We propose a multimodal framework that integrates full-text clinical note embeddings and time-stamped physiological data to jointly model patient outcomes.
Toward a Team of AI-made Scientists for Scientific Discovery from Gene Expression Data
Haoyang Liu, Yijiang Li, Jinglin Jian, Yuxuan Cheng, Jianrong Lu, Shuyi Guo, Jinglei Zhu, Mianchen Zhang, Miantong Zhang, Haohan Wang
ML can discover disease-predictive genes from gene expression data. We introduced the Team of AI-made Scientists (TAIS), a LLM-based framework for automatic streamlining ML analysis. TAIS consists of simulated roles, including a project manager, data engineer, and domain expert.
Selected Projects
The Impact of Productive Failure on Learning Performance and Cognitive Load: Using Hypervideo to Facilitate Online Interactions
Xiaojie Niu, Jingjing Zhang, Kate M. Xu, Xuan Wang
Productive failure is an instructional approach that uses students’ cognitive conflicts to enhance their learning. This experimental study investigated the effect of productive failure in a hypervideo environment.
