Noise Decorrelation as a Hypothetical Mechanism for Phase-Specific Neurometabolic Outcomes in HIV Infection — Code and Data Repository

Computational implementation of "Noise Decorrelation as a Hypothetical Mechanism for Phase-Specific Neurometabolic Outcomes in HIV Infection" This repository contains the Bayesian analysis code, curated neuroimaging datasets, and full reproducibility suite for a study proposing noise decorrelation length (ξ) as a unifying mechanism explaining the 35-year paradox of why neurons survive acute HIV viremia while glia show delayed vulnerability during chronic infection. The framework implements a 5-model Bayesian comparison using PyMC to test whether environmental noise correlation length distinguishes preserved from vulnerable neurometabolic states across HIV infection phases. Key findings: Noise correlation length (ξ) is the dominant predictor of neurometabolic trajectory (ΔWAIC > 15 vs. all alternatives) Acute HIV: neurons maintain short-range noise correlations (ξ ≈ 0.1–1 μm), preserving metabolic homeostasis Chronic HIV: glial populations develop extended noise correlations (ξ ≈ 10–100 μm), destabilizing enzyme kinetics The mechanism resolves the acute preservation / chronic vulnerability paradox without invoking direct viral cytopathology Hierarchical Bayesian inference across 44 individual MRS trajectories with full uncertainty quantification Enzyme kinetics modeling, Tegmark-scale decoherence baselines, and WAIC/LOO model comparison Under review at PLOS Computational Biology (PCOMPBIOL-D-26-00394) Preprint: https://doi.org/10.64898/2026.02.10.703895 (bioRxiv) GitHub: https://github.com/Nyx-Dynamics/noise_decorrelation_hiv Interactive summary (narrated slide deck, mind map, infographic): https://nyxdynamics.org/research/noise-decorrelation/