CellVoyager: AI CompBio agent generates new insights by autonomously analyzing biological data

Area of Science:

• Computational Biology
• Artificial Intelligence in Genomics
• Bioinformatics

Background:

• Modern biology generates high-dimensional data, like single-cell RNA sequencing (scRNA-seq), creating many research hypotheses.
• Analyzing scRNA-seq data is challenging due to time, computational, and expertise requirements.

Purpose of the Study:

• To introduce CellVoyager, an AI agent designed to autonomously generate and execute scRNA-seq analyses.
• To assess CellVoyager's capability in hypothesis generation and data analysis within a Jupyter notebook environment.

Main Methods:

• Development of CellVoyager, an AI agent leveraging large language models for automated scRNA-seq analysis.
• Evaluation of CellVoyager using CellBench, a benchmark comprising 76 published scRNA-seq studies.
• Comparative analysis against GPT-4o and o3-mini for predicting author-conducted analyses.

Main Results:

• CellVoyager outperformed GPT-4o and o3-mini by up to 23% in predicting scRNA-seq analysis choices based on background information.
• In case studies (COVID-19, cell-cell communication, aging), CellVoyager produced novel, expert-validated findings.
• Demonstrated ability to autonomously analyze biological data at scale.

Conclusions:

• CellVoyager significantly accelerates computational biology by automating complex data analysis.
• The AI agent has the potential to uncover novel biological insights previously missed.
• CellVoyager enhances the exploration of vast hypothesis spaces in high-dimensional biological datasets.