A reproducible evidence-to-prior virtual-patient simulation framework for sparse intranasal CNS delivery evidence

This repository contains the reproducibility dataset for the manuscript “ A reproducible evidence-to-prior virtual-patient simulation framework for sparse intranasal CNS delivery evidence ” The dataset supports an in-silico proof-of-concept of a clinically constrained virtual-patient simulation workflow for sparse intranasal CNS delivery evidence. Published aggregate evidence was converted into platform-class modifiers and credibility-based uncertainty widths, then propagated through a normalized semi-mechanistic model separating direct nasal-to-CNS input from systemic redistribution. The package includes Python code, simulation input files, evidence-to-prior mapping files, parameter-prior tables, platform-class modifiers, credibility-based uncertainty modifiers, locked model/metric/constraint definitions, precomputed main simulation outputs, virtual-patient parameter draws, clinical decision outputs, mismatch classifications, representative time-series data, sensitivity outputs, measurement-priority proxy outputs, figure/table source datasets, README files, file manifest, and SHA256 checksums. The dataset contains simulated virtual-patient/scenario data and published-evidence-derived parameterization files only. It contains no individual-level patient data, no human participant data, no electronic health record data, no claims data, no animal experimental data, and no newly collected biological material.