Platform
Our software platform
Connect biological data, explainable analysis and candidate review in a documented scientific workflow. Explore how our internal platform supports your project and what we are developing next.
Find the target
A map of proteins, grouped by similarity.
- Select a protein and see its profile.
- Sites to aim at, and close relatives to avoid.
Build the workflow
Published tools, put together step by step.
- Every step can be changed or swapped.
- Nothing is made until you approve.
| Design | Interface | Fold | PD-L2 risk | Test? |
|---|---|---|---|---|
| bnd-014 | 0.82 | 0.88 | low | ✓ |
| bnd-031 | 0.79 | 0.91 | low | ✓ |
| bnd-022 | 0.77 | 0.85 | low | ✓ |
| bnd-006 | 0.91 | 0.74 | high | – |
Choose what to test
Candidates side by side, with the reasons.
- Scores and counter-screen in one table.
- You decide what goes to the laboratory.
Illustrative preview with illustrative values. Not a live product and not experimental data.
You see why a candidate was chosen, not only that it was
AAanalysis is our open framework for interpretable sequence analysis. It relates predictions to named properties and residue positions, so the scientific basis of a result can be examined and tested.
Published and open. γ-Secretase substrates, AAontology, AAclust, and the AAanalysis package.
The published methods make the basis of a prediction available for review and experimental testing.
The papers behind it →- Which properties, which positionsA prediction comes with the sequence features and the residues behind it, in terms a biologist can read, and not as a score alone.
- A ranking you can questionEach shortlist carries the reasoning for its order, so you can disagree with it before synthesis and not after.
- Questions for the next experimentProperties associated with a prediction suggest what to test. Measurements can then inform the next analysis.
How the work is driven
Our internal software connects tools, project data and scientific review.
Connected tools
MCP provides a common interface for calling software. Workflows connect the tools and data needed for a project.
Methods and records
Written procedures define the task and its checks. A reviewable record links inputs, tool versions, settings, outputs and the decisions behind an analysis.
Human review
Our team reviews scientific decisions and approves laboratory and financial commitments. Additional biosafety controls remain in development.
Read the Field Note on agents, MCP and scientific workflows →
Platform access and deployment
Today
Our internal platform supports commissioned projects, connecting scientific tools, project data and expert review. The MCP server remains under development and is currently used only by our team.
In development
We are developing an integration platform as a service (iPaaS) connecting scientific data, models, workflows and applications. Planned options include InXBio-hosted infrastructure, a combined hardware and software setup, or deployment in your own cloud, with support for your choice of large language models (LLMs).
Tell us the target that is holding your experiment up
You get a straight answer about what is realistic.
Research use only. Scope, criteria and deliverables are agreed per project.