HOMD¶

HOMD provides curated Human Oral Microbiome Database taxonomy, genome metadata, 16S reference sequences, phage, CRISPR, AMR, and phylogeny resources. biodbs supports HOMD as a public download/table catalog.

List FTP Files¶

from biodbs.fetch.HOMD import HOMD_Fetcher

fetcher = HOMD_Fetcher()
files = fetcher.list_ftp()

print(files.names())

To list a subdirectory:

refseq_files = fetcher.list_ftp("16S_rRNA_refseq")

List Batch Downloads¶

downloads = fetcher.list_downloads()
print(downloads.names())

Fetch Tables¶

taxa = fetcher.get_taxon_table()
genomes = fetcher.get_genome_metadata()
crispr = fetcher.get_crispr_table()

df = taxa.as_dataframe()

You can also fetch any tabular HOMD file directly:

table = fetcher.get_table("ftp/some_file.tsv")

Download Files¶

path = fetcher.download_file("ftp/16S_rRNA_refseq/example.fasta", dest="data/homd")

Existing files are kept by default. Use overwrite=True to download again. Large files are streamed to disk.

16S RefSeq (HOMD and MOMD)¶

16S rRNA RefSeq releases are versioned per source. version accepts a release like "15.22" (or "current" for the latest); source is "homd" (default) or "momd" (the mouse database, served from momd.org).

files = fetcher.list_16s_refseq(version="15.22")

# unaligned FASTA + QIIME taxonomy for a pinned HOMD release
homd_fasta = fetcher.download_16s_refseq("data/homd", version="15.22")
homd_taxonomy = fetcher.download_16s_taxonomy("data/homd", version="15.22")

# the same for a MOMD release
momd_fasta = fetcher.download_16s_refseq("data/momd", version="5.1", source="momd")
momd_taxonomy = fetcher.download_16s_taxonomy("data/momd", version="5.1", source="momd")

Without filename, download_16s_refseq selects the unaligned .fasta reference (not the .aligned.fasta/.p9.fasta variants); download_16s_taxonomy selects the .qiime.taxonomy sidecar. Pin an explicit version for reproducible pipelines.

Convenience Functions¶

from biodbs.fetch import homd_get_taxon_table, homd_list_ftp

taxa = homd_get_taxon_table()
files = homd_list_ftp("genomes")

Notes¶

This integration targets public HOMD downloads and reference tables. It does not automate BLAST, login-only pages, or detail-page scraping.