MIDORI2 reference databases

MIDORI2 is a collection of quality-controlled, preformatted reference databases for
taxonomic assignment of eukaryotic mitochondrial sequences, built from GenBank.

This record is release GB273: 15 mitochondrial genes (13 protein-coding plus lrRNA and
srRNA) in 9 output formats (RAW, BLAST, KRAKEN2, DADA2, MOTHUR, QIIME, RDP, SINTAX,
SPINGO), each in nucleotide and amino-acid variants, with and without species-ambiguous
records, and in UNIQ (all unique haplotypes per species) and LONGEST (one sequence per
species) selections. Unpacking the archives gives 1,745 files, 19.6 GB.

MIDORI2_GB273_canonical.zip contains the provenance: the Parquet backbone every
distributed file derives from, the frozen NCBI taxonomy snapshot and its supplement, the
acquisition record, the acceptance-gate evidence, and a record of one bug found and fixed
in the original pipeline.

See README.md for how to choose a file, verify checksums, and cite.