Literature evidence maps
Proposed use: connect a claim to the system studied, the conditions, and supporting sources. Help a team compare evidence while preserving meaningful disagreement.
Give the next breakthrough
an inspectable foundation.
Biological research connects papers, observations, models, and unanswered questions. We are exploring Peel and Lois as a symbolic foundation for that work: trace a claim, examine a connection, and help researchers decide what deserves the next test.
Proposed bioscience applications; no biomedical discovery, clinical validation, or laboratory productivity gain is reported here.
Turn accumulated reading
into knowledge a team can examine.
Pathway and mechanism review
Proposed use: inspect a chain of named relationships. Separate recorded connections from unsupported steps, keeping association distinct from mechanism.
Hypothesis comparison
Proposed use: expose the assumptions and evidence gaps behind competing explanations, helping researchers choose tests that distinguish them.
Continuity across a lab
Proposed use: retain why a study was attempted, what its result supports, and which assumptions did not hold, so the next researcher can revisit the reasoning.
Begin with a question.
Return with evidence.
The workflow we want to evaluate with research partners. This illustration does not run experiments or query a biomedical knowledge base.
- 01
Frame the question
Define a bounded scientific question, the system studied, and the outcome of interest.
- 02
Inspect the evidence
Compare sources and conditions. Keep disagreements and missing evidence visible.
- 03
Propose a hypothesis
Label an explanation as a candidate and expose what it depends on.
- 04
Design the test
A researcher selects an experiment or computational test and records the evaluation criteria.
- 05
Review and retain
Examine the result and its limits. Preserve a version that later work can reassess.
The same words can belong
to different scientific claims.
Species, cell type, timing, and experimental conditions shape what an observation can support. A domain extension needs to carry those qualifiers into comparisons instead of merging statements on surface resemblance.
A missing connection can be valuable
Imagine study A observes a change under condition X, while study B does not under condition Y. A useful research assistant would help the team examine the conditions before deciding the findings conflict. This is an illustrative scenario, not a Peel experimental result.
Structured knowledge needs
domain standards and validation.
The FAIR principles emphasize findability, accessibility, interoperability, and reuse, including detailed provenance. BioModels demonstrates the curation of scientific models against their literature sources. These practices inform our direction; they are not claims of Peel certification or integration.
A bounded pilot.
A clear definition of success.
We want to work with domain teams on a focused question: select shareable evidence, establish expert reference answers and conflict cases, then evaluate whether the system returns dependable paths, appropriate abstentions, and useful material for review.
Research prototype. Not intended for diagnosis, treatment decisions, or replacing experimental validation.
Continue exploringEvidence