Omics-informed vascular digital twins: protocol for a scoping review and framework linking molecular data to mechanistic and hybrid models of large arteries

Objective. To map how multi-gene molecular data have been linked to mechanistic, multiscale and hybrid models of large arteries, and what limits each link. Introduction. Decisions on aneurysm surveillance, restenosis risk and dialysis access depend on how the arterial wall will remodel, which is driven by cellular and molecular processes. Arterial digital twins therefore need molecular data to enter, constrain or test mechanistic models. Existing reviews describe digital-twin architectures, circulatory digital models, vascular ageing or agent-based frameworks; none of those examined reports a systematic search or classifies how molecular data and mechanistic models are linked. Eligibility criteria. Population: large arteries and their cells and molecular data, any species. Concept: a link between molecular-assay data on 3 or more analytes and a mechanistic or hybrid model run in the study, in one of three tiers: data-to-model, co-prediction (mechanistic outputs and molecular data entered jointly in one fitted predictor), and model-guided sampling (model output determines where tissue is sampled). Context: named large-artery beds, generic arteries, grafts, stents and fistulae. Methods. The review will follow JBI methodology and the Preferred Reporting Items for Systematic reviews and Meta-Analyses extension for Scoping Reviews (PRISMA-ScR). After deposit, Europe PMC, PubMed, OpenAlex, arXiv and Semantic Scholar will be searched from 2000, with reference harvesting and citation chasing; recall will be validated on known eligible studies not used to develop the search. Screening and charting will be performed by natural language processing (NLP) software in two independent screener runs with frozen instruction files, calibrated before use against the blind decisions of two independent human auditors and monitored with a control set during use. The auditors will decide every forwarded record and every conflict, audit a random sample of exclusions sized to bound screening sensitivity, and verify every field that feeds a deliverable. Outputs will be a coupling taxonomy, an evidence map, a readiness matrix per omics layer, a resource table with sex and ancestry coverage, and candidate reporting items.