Reproducibility archive: positional-prefix subsampling of archive FASTQ for telomeric-repeat estimation

This archive accompanies the manuscript "Positional-prefix subsampling of archive FASTQ for telomeric-repeat estimation: a preregistered negative benchmark, a rare catastrophic-failure benchmark, and a submission-format marker confirmed prospectively". It contains the protocols, analysis plans and SHA-256 stage locks written before each stage's outcomes existed, frozen cohort manifests (ENA accessions and MD5), per-file results and provenance, frozen analysis outputs, supplementary tables, and the telomere-measurer software with unit tests. REPRODUCE.py (Python 3.11+, standard library only, no network) verifies every file and all 40 lock references and re-runs the frozen analyses of EXP-001, EXP-002, BENCH-001 and EXP-003, comparing each output with the frozen one. No reads from study participants are included; all sequence data are public ENA objects identified by accession. The only read records are three synthetic test fixtures (two FASTQ, one SAM). At release, the submitter free-text column sample_alias and the submitter file names were removed, submitting-centre names were replaced by stable pseudonyms, and abstracts were removed from the literature search exports; RELEASE_TRANSFORMS.json lists the original and released SHA-256 of all 26 changed files. Licences: code (telomere-measurer, REPRODUCE.py and the analysis scripts) is MIT; documentation, protocols, manifests and results written by the author are CC BY 4.0. Third-party metadata (ENA records, bibliographic records) remain under the terms of their sources.