What biological domains does Biomni support?

Biomni supports a wide range of domains including genomics, single-cell RNA-seq, drug discovery (ADMET), GWAS interpretation, proteomics, and clinical phenotype-genotype mapping.

Which LLM models work best with this skill?

While Biomni supports various providers, Anthropic Claude 3.5 Sonnet and Opus are recommended for their superior performance in biological reasoning, task decomposition, and code generation.

Does Biomni require external data downloads?

Yes, upon first use, Biomni automatically downloads approximately 11GB of integrated biomedical databases to your specified local path to facilitate accurate research and knowledge retrieval.

Is code execution safe within the Biomni framework?

Biomni generates and executes Python code to perform complex analyses. For production or sensitive environments, it is highly recommended to run the agent in isolated containers like Docker or virtual machines.

Biomni Biomedical Research Agent

Name: Biomni Biomedical Research Agent
Author: Microck

byMicrock

•

データサイエンスとML

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

概要

Biomni is an open-source autonomous agent framework developed by Stanford SNAP lab, designed to streamline multi-step biological research. It integrates approximately 11GB of curated biomedical databases with LLM reasoning and dynamic code execution to handle specialized tasks like CRISPR screening, single-cell RNA-seq analysis, and ADMET prediction. By bridging the gap between natural language queries and complex computational biology pipelines, Biomni empowers researchers to perform data-driven discovery, variant interpretation, and protocol optimization within a unified, code-integrated environment.

主な機能

Multi-step biological reasoning and autonomous task planning
Built-in evaluation framework for benchmarking agent performance
81 GitHub stars
Integration with 11GB of curated biomedical databases including Ensembl and PDB
Dynamic Python code generation and execution for data pipelines
Cross-domain support for genomics, proteomics, and drug discovery

ユースケース

Predicting ADMET properties and identifying targets for drug discovery projects
Designing genome-wide CRISPR knockout screens and prioritizing gene targets
Analyzing single-cell RNA-seq datasets for cell-type annotation and differential expression

概要

主な機能

Multi-step biological reasoning and autonomous task planning
Built-in evaluation framework for benchmarking agent performance
81 GitHub stars
Integration with 11GB of curated biomedical databases including Ensembl and PDB
Dynamic Python code generation and execution for data pipelines
Cross-domain support for genomics, proteomics, and drug discovery

ユースケース

Predicting ADMET properties and identifying targets for drug discovery projects
Designing genome-wide CRISPR knockout screens and prioritizing gene targets
Analyzing single-cell RNA-seq datasets for cell-type annotation and differential expression