AI Scientist (LLM for Multi-Omics & Precision Medicine)

Job Title:  AI Scientist (LLM for Multi-Omics & Precision Medicine)

About the Lab

The Dong Lab (http://www.donglab.org) at Yale School of Medicine is seeking a highly motivated AI-focused scientist to develop next-generation large language model (LLM) approaches for gene regulation and precision medicine.

Our lab leads the NEUROMICS platform, a large-scale effort integrating:

• Thousands of human brain samples with single-cell and bulk multi-omics 
• Multi-layer regulatory features (splicing, polyadenylation, circRNA, RNA modification, eRNAs, xQTLs, etc.) 
• Millions of longitudinal EHR records (e.g., COSMOS and other real-world datasets) 

We have built a robust data foundation and standardized pipelines. The next phase is to leverage AI/LLMs to transform these datasets into predictive and mechanistic models of human disease.

Position Overview

We are looking for a creative and driven researcher to develop and apply LLM-based models to large-scale biological and clinical datasets.

This role focuses on building foundation models for gene regulation and disease, with opportunities to connect computational predictions to organoid-based experimental validation, forming a closed-loop discovery system.

Key Responsibilities

You will lead or contribute to one or more of the following directions:

1. Virtual Cell Modeling

• Develop LLM-based or foundation models (e.g., extending GeneFormer-like architectures) 
• Model gene regulatory programs across cell types and conditions 
• Integrate multi-modal features beyond gene expression (e.g., splicing, RNA processing, regulatory elements) 

2. Disease Modeling

• Build predictive models for neurodegenerative diseases (PD, ALS, AD) 
• Integrate multi-cohort omics data to model disease onset, progression, and heterogeneity 
• Link molecular states to clinical phenotypes using large-scale EHR data 

3. Therapeutic Discovery

• Perform AI-driven drug repurposing using real-world data 
• Infer therapeutic targets via gene–drug interaction resources (e.g., CMAP-like datasets) 
• Develop models for drug–compound similarity and novel therapeutic inference 

4. Model Development & Engineering

• Fine-tune or pretrain large-scale models on multi-omics datasets 
• Design scalable pipelines for training and evaluation 
• Contribute to open-source tools and reproducible workflows 

Required Qualifications

• Strong coding skills (Python required; experience with ML frameworks such as PyTorch, JAX, or TensorFlow) 
• Demonstrated experience or strong interest in AI/ML modeling (especially deep learning or LLMs) 
• Ability to work with large-scale datasets (cloud/HPC experience preferred) 
• At least one first-author publication or major project (paper, preprint, or equivalent work) 
• Strong problem-solving skills and ability to work independently 

Preferred Qualifications (not required)

• Experience with LLMs or foundation models (e.g., transformers, generative models) 
• Exposure to computational biology, genomics, or biomedical data 
• Familiarity with gene regulation concepts 
• Experience with sequencing data analysis 
• Active GitHub contributions or open-source involvement 
• Experience with AI-assisted coding tools (e.g., Claude Code, Copilot) 
• Background in multi-modal data integration 

Who Should Apply

We welcome candidates from diverse backgrounds, including:

• Computer science / AI / data science 
• Physics / math / engineering 
• Computational biology / bioinformatics 

A formal background in molecular biology is not required, but curiosity about biological systems is important.

What We Offer

• Access to unique, large-scale multi-omics and EHR datasets
• Strong computational infrastructure (GPU clusters, cloud resources) 
• Close integration with experimental platforms (organoids, in vivo models) 
• Highly collaborative environment across Yale and external partners 
• Opportunity to lead high-impact projects at the intersection of AI and medicine 

Application

Please send:

• CV 
• Brief statement of research interests 
• GitHub (if available) 
• Representative work (papers, preprints, or projects) 

to: xianjun.dong@yale.edu