Claude Science can run your protein engineering analysis. Whether you can trust the results depends on the data model it’s connected to.
In 30 minutes we connect Claude Science to LabKey Biologics LIMS over MCP and put it to work on a real problem: 12 sequence variants, aligned to the parent, and screened for binding and neutralization.
We’ll walk through how Biologics LIMS makes this analysis possible in Claude Science: molecules, sequences, samples, assay results, and inventory, each registered with a stable ID and connected by modeled relationships. You’ll see which queries the agent runs, which relationships it follows, and how every figure traces back to Biologics LIMS as the system of record.
What you’ll see in Claude Science and Biologics LIMS:
- A sequence variant set aligned against its parent in Claude Science, with substitutions resolved from the bioregistry rather than from a FASTA file someone maintained by hand.
- Binding and neutralization data pulled by molecule, already joined to the samples and constructs that produced it, then plotted without a manual merge step.
- Inventory queried in the same session, confirming which variants have enough material and where the vials sit before anyone commits to a follow-up experiment.
- Analyses and a written lead-candidate rationale posted back into the ELN through the LabKey API, closing data, analysis, and documented conclusion inside one system.