Long COVID Mechanistic Pathway Atlas v3.2

Long COVID affects an estimated 65–200 million people worldwide and has generated a mechanistic literature with more than five thousand peer-reviewed primary papers across more than sixty distinct candidate drivers. The v3.2 atlas applies an eleven-technique analytical stack to 5,101 papers and 222 mechanism link cards to derive a hierarchical-DAG-with-cycles architecture: three causal roots (R1 viral persistence, R2 microbiome dysbiosis, R3 endothelial coagulopathy), ten amplification hubs (H2–H13), and twenty-seven mediators converging onto nine symptom domains. The principal findings include a 156-pair inter-root directed-pair verdict matrix — the first comprehensive structural map of its kind for any mechanistic-disease literature — a dual-version v3.2 / v3.2.1 evidence-tier audit that distinguishes conceptual-coherence curation from strict peer-review evidence grading, a 4-universal-fail field diagnosis showing that four chronic gates (effect-size CI, prospective design, ROBINS-I-passing, PI-independence) fail at field-level frequencies of 59–96% across all 13 chains, the identification of SIM01 → X.Dysbiosis as the sole Pearl-defensible therapeutic claim (composite priority score 9.00 vs runner-up 1.96), and a seven-item v3.3 RCT priority queue ranked by composite-influence weight. The atlas resolves the four-year theoretical dispute between the Groysman 2024 flat-network frame and the Proal–VanElzakker 2021 triadic-root frame by absorbing both as projections of the hierarchical-DAG-with-cycles architecture.