Lab profile
About Sunlab
Sunlab advances artificial intelligence for healthcare across drug discovery, clinical trial optimization, computational phenotyping, clinical predictive modeling, treatment recommendation, and health monitoring. The lab focuses on building AI for real-world healthcare, clinical trials, and therapeutic science with clear paths to clinical and scientific translation.
Clinical prediction & decision supportLarge language models for medicineAI for drug discoveryBiosignals: EEG, ECG and sleepSynthetic health data & privacyDesigning trials and predicting their outcomesMatching patients to the right trialsSynthesizing clinical evidence with LLM agents
01 · Research focus
Research areas and highlights
Topics and recent work as stated on the lab’s own site.
- RETAIN: An Interpretable Predictive Model for Healthcare using Reverse Time Attention MechanismA two-level attention model that reads a patient's visits in reverse time order, highlighting which past visits and diagnoses drove each prediction.
- GraphCare: Enhancing Healthcare Predictions with Personalized Knowledge GraphsBuilds a personalized knowledge graph for each patient using large language models and biomedical knowledge bases, then reasons over it with a graph neural network.
- DRG-LLaMA: tuning LLaMA model to predict diagnosis-related group for hospitalized patientsFine-tunes the LLaMA language model on discharge summaries to predict each hospital stay's DRG code directly from text.
- Therapeutics Data Commons: Machine Learning Datasets and Tasks for Drug Discovery and DevelopmentAn open platform and Python library unifying 66 AI-ready datasets across 22 learning tasks for therapeutic machine learning.
- MolTrans: Molecular Interaction Transformer for drug-target interaction predictionBreaks drugs and proteins into substructures, learns from large unlabeled molecular databases, and models their pairwise interactions with a transformer.
- BIOT: Biosignal Transformer for Cross-data Learning in the WildTurns any biosignal into a uniform sequence of tokens, letting one transformer pre-train across EEG, ECG and activity datasets with mismatched formats.
- HINT: Hierarchical interaction network for clinical-trial-outcome predictionsA graph neural network that combines drug molecules, target disease, and eligibility criteria to predict whether a trial succeeds.
- AutoTrial: Prompting Language Models for Clinical Trial DesignA language model that drafts eligibility criteria from a trial's setup and similar past trials, explaining its reasoning step by step.
- COMPOSE: Cross-Modal Pseudo-Siamese Network for Patient Trial MatchingA two-path neural network that encodes eligibility criteria and longitudinal health records for patient-trial matching.
- TrialBench: Multi-Modal AI-Ready Datasets for Clinical Trial PredictionA suite of curated, multi-modal datasets from ClinicalTrials.gov covering tasks such as trial duration, dropout, adverse events, approval, failure reasons, and dose finding.
02 · People
Leadership and researchers
People listed on the lab’s team pages, by role.
Jimeng SunDirector
Chufan (Andy) GaoPhD student
Patrick JiangPhD student
Siddhartha LaghuvarapuPhD studentLinkedIn ↗
Jiacheng LinPhD student
Jathurshan PradeepkumarPhD student
Hanyin WangPhD student
John WuPhD student
03 · Updates
Updates and lab news
News and events from the lab’s own site.
- TrialMind published in npj Digital MedicinePublicationdoi.org
- LEADS published in Nature CommunicationsPublicationdoi.org
- PyHealth 2.0 releasedAnnouncementpyhealth.dev
- DeepEvidence published in Nature Machine IntelligencePublicationdoi.org
- TrialPanorama published in npj Digital MedicinePublicationdoi.org
04 · Official presence
Lab sources and channels
The lab’s own pages and accounts.