Skip to content
An open scientific journal beside a glowing protein structure

Publications

Publications

Advancing open science. Our methods, definitions and results — written up so others can check our work.

Topic

6 results

  • An explainable evidence engine for ranking gene–disease targetsPlaceholder

    ARYAN AI Technical note (2026) · Target discovery · Forthcoming

    How genetic association, expression, pathway context and tractability are combined into a transparent target score — and how we audit it.

  • Multi-omics integration for target identification in inflammatory diseasePlaceholder

    ARYAN AI Technical note (2026) · Omics · Forthcoming

    A reproducible pipeline for bulk and single-cell RNA-seq with batch correction, pathway scoring and links to candidate targets.

  • Structure-aware generative chemistry with synthesis-aware scoringPlaceholder

    ARYAN AI White paper (2026) · Generative chemistry · Available on request

    Design principles behind our generative models: pocket conditioning, multi-parameter optimisation and retrosynthetic feasibility.

  • Active learning to choose the next experimentPlaceholder

    ARYAN AI Technical note (2026) · Methods · Forthcoming

    Using model uncertainty to select the compounds that teach the model most per assay, and the effect on design-cycle efficiency.

  • Early developability prediction for therapeutic antibodiesPlaceholder

    ARYAN AI Technical note (2026) · Biologics · Forthcoming

    Sequence- and structure-based flags for aggregation, polyspecificity and immunogenicity risk applied before expression.

  • Designing siRNA with learned potency and off-target modelsPlaceholder

    ARYAN AI Poster (2026) · Methods · Forthcoming

    Ranking siRNA candidates by predicted knock-down efficiency while penalising seed-region off-target matches.

ARYAN AI has not yet published in peer-reviewed journals. Items above are planned company technical notes; citations will be added on publication.