typeaudit v1.4.0 — metadata audit of three fungal genera
收藏资源简介:
Genus-level metadata audit of three fungal genera (Colletotrichum, Trichoderma,Diaporthe). For 1,360 accepted binomials we cross-check the strain numbers andtype-material numbers declared in GenBank and NCBI Taxonomy against the publiccatalogues of three collections (CBS, ICMP, CGMCC) and against MycoBank's currentnames. Of 11,116 number-level cells, 3,040 (27.3%) can be verified in a collectioncatalogue; within those, the number-level discrepancy rate is 13.3% (95% CI12.1–14.5%), an upper bound. This archive contains the derived tables (DATA/), code that recomputes everyheadline number offline (CODE/reproduce.py, Python standard library only), acolumn dictionary with the flag glossary (DOCS/column_dictionary.md), a versionmap (DOCS/version_correspondence.md) and a note on third-party sources(DOCS/third_party_data.md). Run `python CODE/reproduce.py` to verify the numbers What changed in v1.4.0 (2026-10-29) - Self-contained English package: 3.7 MB of derived tables, all column headers in English, plus DOCS/column_dictionary.md glossing every column and flag. - CODE/reproduce.py recomputes every headline number of the paper offline (11,116 cells / 3,040 verifiable / 559 / 403 / 27.3% / 13.3%) from the shipped tables, using only the Python standard library. - The Windows launcher (.exe, 14 MB) has been removed; the pipeline itself lives in the GitHub repository. - Corrected the cell total from 11,103 to 11,116 (the per-genus tables were regenerated on 2026-10-24; see DOCS/version_correspondence.md). - Third-party sources (GenBank/NCBI, MycoBank, CBS/ICMP/CGMCC) are not redistributed; DOCS/third_party_data.md says how to refetch them.



