Biomni

关于

Biomni is an open-source biomedical AI agent framework from Stanford's SNAP lab designed to automate complex research across the life sciences. It provides a unified interface for conducting multi-step biological reasoning, generating and executing analysis code, and accessing an integrated 11GB database of biomedical literature and datasets. Whether designing CRISPR screens, interpreting GWAS results, or predicting drug ADMET properties, Biomni empowers researchers to bridge the gap between high-level biological inquiries and technical data processing through intelligent task decomposition and autonomous execution.

主要功能

• Dynamic code generation and execution for biomedical data analysis
• Autonomous task decomposition and multi-step biological reasoning
• 16 GitHub stars
• Access to 11GB of integrated knowledge sources like Ensembl, UniProt, and PubMed
• Built-in evaluation framework to benchmark agent performance
• Cross-domain problem solving across genomics, proteomics, and clinical tasks

使用场景

• Predicting ADMET properties and optimizing compounds for drug discovery
• Designing CRISPR knockout screens and prioritizing genes based on pathway relevance
• Analyzing single-cell RNA-seq data for cell type annotation and clustering