INTRO

Steven asked me (GitHub Issue) to generate a DIAMOND BLAST database for use with his class, as well as our lab, since we didn’t currently have one set up.

MATERIALS & METHODS

Downloaded NCBI NR FastA from here:

https://ftp.ncbi.nlm.nih.gov/blast/db/FASTA/nr.gz
Confirmed MD5 sum was okay (not shown)
Renamed the unzipped FastA to match download date: ncbi-nr-20250429.fasta

Caution

This is a huge FastA file. Unzipped, it will be ~350GB in size!

Downloaded NCBI Taxonomy info from here:

https://ftp.ncbi.nlm.nih.gov/pub/taxonomy/taxdmp.zip
Confirmed MD5 sum was okay (not shown)

Before creating the DIAMOND BLAST database, I had to manipulate some of the NCBI files first.

Had to replace superkingdom designation in NCBI NR node.dmp, as this is a recent change at NCBI and DIAMOND does not yet have a release to handle this change in nomenclature.

sed -i 's/domain/superkingdom/g' nodes.dmp

And, I actually had to replace all the new ranks (“domain”, “realm”, “acellular root”, and “cellular root”) in nodes.dmp and names.dmp with “superkingdom” (figured out due to this comment in this GitHub issue).

Make the Database

Ran this command to make the database:

/home/shared/diamond-2.1.8 makedb \
--in ncbi-nr-20250429.fasta \
--db ncbi-nr-20250429 \
--taxonmap prot.accession2taxid \
--taxonnodes nodes.dmp \
--taxonnames names.dmp \
--threads 40

The --db ncbi-nr-20250429 simply specifies the output file name to be used for the resulting database.

RESULTS

The final database is located here:

/home/shared/16TB_HDD_01/sam/databases/blastdbs/ncbi-nr-20250429.dmnd

The database is 349GB!

SUMMARY

Overall, the entire process of downloading, unzipping, and creating the database took nearly 6hrs.