These files provide the NCBI taxonomic data for a cleaned up subset of fasta sequences from the Genometa paper by Davenport et al (PloS one 7.8 (2012): e41224 DOI: 10.1371/journal.pone.0041224).
The text file "gi-taxonomy-maps.clean.txt" can be readily parsed, allowing the GI numbers in the fasta headers from Genometa to be mapped onto NCBI taxonomic information. This allows the species, genus etc of the genome sequences to be retrieved.
There are also two fasta files based on the related Genometa fasta files, providing a set of species, and a set of strains of those species. These files are the compressed "allgenomes_april_2012_v6_one_per_species.fa.cleaned" and "allgenomes_subset.list.fa.cleaned" respectively. The fasta headers contain the GI numbers and name of the genome sequence.
The taxonomy file allows the GI numbers in each fasta file to be mapped to each other (i.e. strains to species), or to other taxonomic levels.