Does Biomni require a specific LLM to function?

Biomni is flexible and supports various providers including Anthropic Claude (recommended), OpenAI, Azure, Google Gemini, and AWS Bedrock.

Is it safe to run Biomni locally?

Because Biomni executes LLM-generated code with system privileges, it is highly recommended to run the agent in isolated environments like Docker or virtual machines.

Does the skill include biological data?

Biomni integrates an ~11GB biomedical data lake containing gene, protein, clinical, and pathway databases which it downloads automatically on first use.

Can Biomni analyze my own datasets?

Yes, Biomni can autonomously generate and execute analysis code to process local data files, such as .h5ad files for single-cell RNA-seq or GWAS summary statistics.

Biomni is an autonomous AI agent framework developed by Stanford SNAP lab specifically for executing complex research tasks across biomedical domains like genomics and drug discovery.

Biomni Biomedical AI Agent

Name: Biomni Biomedical AI Agent
Author: Microck

byMicrock

•

110

データサイエンスとML

Automates complex biomedical research tasks including genomics, drug discovery, and clinical analysis using autonomous AI reasoning and code execution.

概要

Biomni is an open-source framework from Stanford's SNAP lab designed to function as an autonomous biomedical AI agent within Claude. It bridges the gap between high-level research questions and technical execution by combining advanced LLM reasoning with dynamic code generation and access to a massive 11GB integrated biomedical database. Whether designing CRISPR screens, analyzing single-cell RNA-seq data, or predicting drug ADMET properties, Biomni handles task decomposition, data retrieval, and complex analysis pipelines to accelerate discovery in molecular biology, proteomics, and clinical genomics.

主な機能

Autonomous multi-step biological reasoning and research task planning
Cross-domain support for genomics, drug discovery, and clinical tasks
110 GitHub stars
Integrated access to 11GB of curated biomedical databases and literature
Dynamic code generation and execution for custom data analysis pipelines
Built-in evaluation framework for benchmarking biomedical agent performance

ユースケース

End-to-end single-cell RNA-seq processing including clustering and cell type annotation
Predicting ADMET properties and toxicity for drug discovery candidate optimization
Designing CRISPR knockout screens with automated gene prioritization and pathway analysis

概要

主な機能

Autonomous multi-step biological reasoning and research task planning
Cross-domain support for genomics, drug discovery, and clinical tasks
110 GitHub stars
Integrated access to 11GB of curated biomedical databases and literature
Dynamic code generation and execution for custom data analysis pipelines
Built-in evaluation framework for benchmarking biomedical agent performance

ユースケース

End-to-end single-cell RNA-seq processing including clustering and cell type annotation
Predicting ADMET properties and toxicity for drug discovery candidate optimization
Designing CRISPR knockout screens with automated gene prioritization and pathway analysis