Results

A Genomarker result is not just a number — it is a bundle of evidence about how that number was produced. This is what every result carries.

Reproducibility Receipt

Every analysis emits a signed Reproducibility Receipt: pinned seeds, software versions, a dataset fingerprint, and a canonical RFC 8785 hash with SHA-256, under frozen schema 1.11.0. Published receipts live at public verify links, so anyone can check exactly which data, code, and parameters produced a result. Receipts attest provenance.

Reproducibility Score

Each inferential receipt also carries a Reproducibility Score — 0–100 with a letter grade, composed over 12 axes of rigor including calibration, environment capture, pre-registration, and reporting completeness — with an embeddable SVG badge for papers, posters, and lab pages.

Reporting checklists

The platform auto-emits the reporting checklists journals increasingly require — TRIPOD+AI 2024, REMARK 2005, and STARD-AI 2025 — generated from the analysis record itself, not from the author’s memory.

Audit-Grade Dossier

For studies that need to withstand formal review, the platform assembles an Audit-Grade Dossier: typed Claims, the receipts that support them, figures, and Methods — minted with a Zenodo DOI so the whole bundle is citable.

Methods-Diff

Receipts can be diffed — against each other, or against the methods section of a published paper retrieved via Europe PMC, Crossref, or arXiv — to show precisely what changed between two analyses.

Publications

A curated list of the publications supporting this work — including Dr. Sayad’s books and papers on real-time machine learning and data mining — is maintained on the About page and on his ResearchGate profile.