Claude Life Sciences Marketplace

Why it's ranked here

The collection is unusually practical because it pairs service connections with executable analysis workflows. The strongest evidence is its single-cell tooling: validation, configurable filtering, model training, clustering, integration, label transfer, plots, and saved results form a coherent path. Its value still depends heavily on external accounts and local scientific software.

What's good

The single-cell workflows expose meaningful controls instead of hiding scientific choices. Users can tune quality thresholds, highly variable gene selection, latent dimensions, training epochs, clustering resolution, and prediction confidence. Validation checks raw counts, missing metadata, invalid values, sparsity, batch sizes, rare labels, and gene selection before expensive training begins.

Tradeoffs

Most service connections require authentication, and some require paid or institutional access. The 10x Genomics connection only helps users who already have cloud analysis data. Nextflow workflows require Docker and Nextflow locally. Provider support remains separate, and tenant-specific service URLs are unsupported, which caused the Benchling integration to be removed.

How to use it well

It suits life scientists already working in Claude Code who want literature access and repeatable omics workflows close to their research conversation. Start with one narrowly relevant plugin, validate inputs, inspect generated plots, and retain explicit parameters for reproducibility. Treat the single-cell quality-control workflow as preprocessing, not a complete cleanup suite: ambient RNA correction and doublet detection remain separate next steps.

Observed

Install surface

Claude Code marketplace commands install individual MCP servers and skills.

Interfaces

Remote MCP, local MCPB, Claude Code skills, and Python command-line scripts.

Authentication

PubMed requires none; other listed services require free, paid, or institutional accounts.

Platform requirements

The Nextflow workflow requires Docker and Nextflow installed locally.

Implementation language

The supplied analysis and quality-control scripts are Python.

License

Individual MCP servers use their providers' licenses and terms of service.