Biomni FAQs

Question 1

What is Biomni?

Accepted Answer

Biomni is a Databricks application that exposes 18 bioinformatics command-line interface (CLI) tools via the Model Context Protocol (MCP), enabling AI agents on Databricks (like Genie Code, Claude, or custom agents) to seamlessly access complex bioinformatics functionalities.

Question 2

How does Biomni integrate with Databricks AI agents?

Accepted Answer

Biomni uses the Model Context Protocol (MCP) for communication, allowing Databricks AI agents to call its tools using natural language. It integrates directly with Genie Code and supports custom agents, secured via OAuth authentication.

Question 3

What are Biomni's execution tiers and why are they important?

Accepted Answer

Biomni offers four execution tiers: In-App (Python packages), Spark (distributed with Glow Pipe), Cluster (file-based genomics tools), and GPU (deep learning tools like nnUNet). These tiers optimize performance and resource usage for different tool requirements.

Question 4

Is custom cluster configuration required to use Biomni?

Accepted Answer

Tier 1 (In-App) tools work immediately without extra setup. For Tier 2, 3, and 4 tools that require distributed or specialized computing, you will need to configure Spark, standard, or GPU clusters and potentially set up Unity Catalog Volumes for data I/O.

Question 5

Which types of bioinformatics tasks can Biomni assist with?

Accepted Answer

Biomni covers a wide range of bioinformatics tasks, including RNA secondary structure prediction, plasmid annotation, protein conservation analysis, NCBI BLAST searches, genomics pipelines (BWA, GATK), medical imaging, and molecular docking, among others.

Biomni

Biomni

主要功能

使用案例

主要功能

使用案例