PerioAlzheimer-GPVS: Computational Data and Python Scripts for Gingipain-Sensitive Proteolytic Vulnerability Mapping

PerioAlzheimer-GPVS: Computational Data and Python Scripts for Gingipain-Sensitive Proteolytic Vulnerability Mapping This repository contains computational scripts, processed datasets, and supplementary materials supporting the study entitled "Computational mapping of gingipain-sensitive proteolytic vulnerability at the periodontitis–Alzheimer disease interface." The study evaluates the potential susceptibility of 24 human proteins associated with periodontal inflammation and Alzheimer's disease to gingipain-mediated proteolysis by Porphyromonas gingivalis. The computational framework integrates three complementary dimensions: (i) sequence-based susceptibility determined from lysine and arginine candidate positions, (ii) structural and disorder-aware plausibility derived from AlphaFold-predicted protein models, and (iii) functional relevance based on proximity to annotated protein features. These components are integrated into a Gingipain Proteolytic Vulnerability Score (GPVS), with ranking robustness evaluated using z-score, percentile-based, and geometric aggregation approaches. The repository also documents exploratory experimental concordance mapping of the tau-441 isoform and sensitivity analyses of the structural classification thresholds. The materials are intended to facilitate computational reproducibility, independent verification, and further investigation of candidate molecular targets at the periodontitis–Alzheimer disease interface. GPVS is a hypothesis-generating prioritization measure and does not establish experimentally confirmed gingipain cleavage or a causal relationship between periodontal disease and Alzheimer's disease.