Gene-disease associations
The work focuses on separating established biology from emerging and exploratory candidate genes.
Research-use only
Evidence-ranked disease-gene hypothesis research.
TensorCortex is a research-use methods project exploring how public biomedical evidence can be structured, scored, and reviewed before disease-gene hypotheses are surfaced.
Public biomedical sources only
No medical or treatment guidance
Traceable evidence paths
Review gates before publication
Methods surface
From public evidence to reviewed hypotheses
A restrained workflow for turning source metadata, normalized evidence, and review outcomes into research-priority material.
01Source metadatatraceable
02Identity normalizationauditable
03Candidate scoringbounded
04Scientific reviewrequired
Sources
Rows
Candidates
Review
Research boundary
The project studies how hypotheses can be made more inspectable. It does not provide diagnosis, treatment guidance, patient risk prediction, or clinical decision support.
Evidence traceability
Candidate claims need visible source context, normalization history, and a path back to disease biology.
Method controls
Controls, decoys, uncertainty checks, and review gates are treated as part of the research output.
Method principles
Evidence before explanation.
Language can make a hypothesis readable, but it should not invent the hypothesis. The research boundary is evidence first, explanation second.
Structured sources
Source metadata, identifiers, licensing, and provenance stay explicit instead of becoming hidden prompt context.
Normalization boundaries
Evidence rows are normalized before scoring so source-native statistics remain visible and auditable.
Scoring boundaries
Scores rank research candidates; they are not diagnostic, clinical, or causal claims.
Uncertainty review
Exploratory hypotheses require caveats, failure modes, and suggested validation before public surfacing.
Publication status
Public research artifacts are not available yet.
Public pages will only be published after evidence review, uncertainty review, research-use language checks, and source-boundary checks. The current site is here to establish the methods surface without implying medical, diagnostic, or treatment use.
research use onlyevidence reviewpublication gated