Research evidence / Results you can interrogate

See the decisions the science changed.

A corrected geometry reopened a bivalent design. Kinetic models favored a simpler covalent candidate. The records below show how BioTwin changed a research decision and what evidence comes next.

The Lp(a) design decision

The Lp(a) case follows a completed computational design progression. A bivalent architecture was compared with a simpler monovalent covalent strategy. The following is a case-level summary of the recorded work, with no claim of experimental activity.

  • 1. Correct the geometry · A domain-center distance had ruled out a bivalent design. A pocket-to-attachment analysis changed that geometric conclusion.
  • 2. Compare the mechanisms · Kinetic and regimen models supported continuing the simpler covalent design while retaining the bivalent option.
  • 3. Define physical validation · The outcome was a computational candidate program with alternatives and a physical validation plan. Download case summary (JSON)

Examine the atlas and the underlying representations

The E3 atlas documents the breadth of one search. The 8DB examples examine how scientific relationships and alternatives can be represented and queried.

  • E3: computational scope made explicit · The published atlas distinguishes screened ligases, included adaptors, prioritized ligases and recruiter-anchor hypotheses. None of those counts is a binding or degradation result. Inspect the atlas scope
  • 8DB: preserve the reason for the next step · Native scientific representations and declared dependencies make an investigation available for comparison and follow-up. Planning suggestions and evidence-frontier comparisons have separate demonstrations. Examine the foundation

What each kind of result establishes

A successful software operation, a computational comparison and a biological experiment answer different questions. Capability pages describe which kind of work exists and where further evidence is needed.

  • Implemented operation · The method executes within a specified computational scope.
  • Computational result · A model, comparison or candidate supports a bounded scientific conclusion.
  • Experimental result · A physical measurement tests a molecular or biological claim.
  • Clinical outcome · A patient study evaluates a clinical question under its protocol.

When a result narrows the claim

A useful method must reveal where it stops helping. The topological pocket work localized known sites, but a broader apo-only test did not establish a general discovery advantage. In Lp(a), a kinetic model showed why reversible capping alone would not sustain lowering. These findings guide the next test and keep a result from being applied beyond its scope.

Compare against your existing workflow

A prospective evaluation should define the target, input data, existing workflow and decision before work begins. Compare useful hypotheses, evidence quality, the next discriminating experiment and research effort under the same conditions. Cost or speed improvements should be measured in that evaluation.

Review the supporting research

Start with a nonconfidential scientific question. We can discuss the supporting methods and scope an appropriate review of research material, including access and confidentiality requirements.

Discuss a focused scientific evaluation

Tell us the target, the constraint and the decision your team needs to make.