Mutation by Natural Dominance: source-audited evidence, competing-risk diagnostics, and limits of mechanism identification

This exploratory technical report presents source-audited evidence, mathematical derivations, and reproducible software for evaluating Mutation by Natural Dominance (TMD) hypotheses. It is research extension version 0.1.0, distinct from the archived software v2.6.5.

Main findings and methods

A context-stratified diagnostic tests changing route fractions across prespecified first-successful-arrival time intervals under independent right censoring. An explicit mathematical construction shows that common gamma frailty, independent route-specific gamma rates, and deterministic decreasing hazards can produce exactly the same first-event joint distribution. A mutation-supply-adjusted WS restriction clarifies comparisons when baselines vary across contexts. These derivations apply standard probability; external mathematical novelty is not claimed.

A secondary transcription and analysis of Leehan and Nicholson (2021), Table 1 preserves three experimental blocks and 111 resistant-isolate observations. Context-specific predictions gain 13.126 bits over a pooled model across held-out blocks. The analysis is exploratory, not preregistered, and reports no significance or causal claim. The source audit also corrects the use of pathway-isolated WS sample totals as competing-route winner frequencies and distinguishes duplicate panels and synthetic timings from independent biological observations.

Software verification

All 16 tests passed on Python 3.9.6 and 3.12.14. In selected simulations with 200 datasets per scenario, null rejection rates were 12/200 and 9/200 at a 5% threshold; a deliberately strong switching alternative was detected in 200/200. These are synthetic software evaluations, not biological measurements or universal calibration. The public GitHub snapshot was verified byte-for-byte.

In More Basic Terms

Evolution has several possible routes to a useful change. Counting which route appears most often does not automatically explain why it succeeds. This report checks the origins of older evidence and provides a new test: do early and late successful changes use the same mix of routes within the same conditions? Different mechanisms can create identical timing patterns, so independent mutation-rate and performance measurements are needed. The aim is to make predictions that can fail fairly on new data.

Scope, sources, and reuse

No new laboratory experiment, independent biological validation, unique mechanism identification, or new fundamental physical law is reported. The rpoB observations belong to the cited original researchers; their numerical table is transcribed with attribution and cell-level provenance. AI assistance was used in preparing the analysis, code, and documentation. Neither EDISON grant 675419 nor MOSBRI grant 101004806 funded TMD.

The report and new documentation are CC BY 4.0; software is MIT. Third-party observations and inherited files retain their specific attribution and notices. The accompanying ZIP is the unchanged GitHub research-extension bundle and contains code, tests, data provenance, literature review, mathematical notes, and a prospective empirical test plan.

GitHub research release: https://github.com/maldonado-research/TMD/releases/tag/v0.1.0
Earlier archived software: https://zenodo.org/records/22398093
Primary data source: https://doi.org/10.1128/AEM.01237-21
WS source: https://doi.org/10.7554/eLife.38822
Competing-risk background: https://doi.org/10.1073/pnas.72.1.20