Biodius

Biodius / Approach

Make the evidence earn the next step.

A convincing biological story is a starting point. The review asks which causal statement has earned the next expensive step.

The falsification chain

  1. Claim / target thesis
  2. Source identity + provenance
  3. Human-genetics / causal-gene evidence where applicable
  4. Study/cohort lineage collapse
  5. Contradictions, nulls and failed programmes
  6. Causal-chain audit
  7. Novelty/prior-art attack
  8. Independent synthetic expert challenge
  9. Primary-source adjudication
  10. Cheapest discriminating next evidence
  • ADVANCE
  • REWORK
  • STOP
  • NOT_EVALUABLE
A schematic: freeze the question, challenge the evidence, adjudicate against primary sources, then record ADVANCE, REWORK, STOP or NOT_EVALUABLE. 01 / FreezeQuestion + stop rule02 / ChallengeEvidence + causal chain03 / AdjudicatePrimary sourcesRecord what survives — or cannot be decided.
Method schematic · An explicit decision, including NOT_EVALUABLE.

Why ordinary AI research is not enough

AI research tools make it fast and cheap to assemble a convincing biological story. Throughput is not truth: a well-written summary can turn “plausible and not yet disproven” into confidence. A target decision needs an attempt to break the story before money is committed.

How AI is used — and why the output should be checked

AI research agents do the broad evidence gathering, and independent AI reviewers are tasked with falsifying the case rather than improving it. Their challenges are then adjudicated against primary sources, and the analysis rules are fixed before the decisive evidence is read. We do not claim the AI is unbiased; the workflow is designed so that a weak claim has to survive more than one attempt to break it.

The method has not been externally validated. Its current commercial hypothesis is usefulness as scientific diligence before a costly downstream decision.

Experimental by design. Bounded in scope.

Biodius is experimental scientific diligence / falsification tooling. It is not a clinical system, a therapeutic recommendation engine, a validated drug-discovery platform, a substitute for qualified scientific judgment, or proof that any biological target is causal, safe or effective.

Biodius does not provide clinical advice, validate safety or efficacy, perform wet-lab work, or replace legal, patent or regulatory review.

What would change your decision?

Send one target + indication + rationale, and the next decision you are considering.

Start a Target ChallengeRead the preprint →