This is a test version of Biostars. For the public version, visit https://www.biostars.org.
what's sequences are in the precomputed human_genomic database

I recently installed blast and downloaded the precomputed human_genomic.*tar.gz database available here:

ftp://ftp.ncbi.nlm.nih.gov/blast/db/

I tested my installation with the following fasta file:

cat test_query.fa 
>chr13:83987454-83987503
GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT

when I blast against my local database I see the primary assembly but also many additional hits:

>NC_000013.11 Homo sapiens chromosome 13, GRCh38.p7 Primary Assembly <- matches my test query
Query  1         GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT  50
Sbjct  83987454  GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT  83987503

>NT_024524.15 Homo sapiens chromosome 13 genomic scaffold, GRCh38.p7 Primary 
Query  1         GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT  50
Sbjct  65579348  GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT  65579397

>GL583019.1 Homo sapiens unplaced genomic scaffold scaffold_39, whole genome 
Query  1       GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT  50
Sbjct  731735  GCTGGGTGGTCAGCGCTGGTTCCATGGGCAGTAATGATTTCCTCTGTTTT  731686

>Lots more results...

My question is what is the source of all additional sequences that this blast database uses?

I have looked at the README (available at ftp://ftp.ncbi.nlm.nih.gov/blast/db/README) but the information there is not very thorough. Is there a complete list of what's in this database? Thanks!

blast

1 answer

See if this helps:

Capture

this is better than the README file but when I use blast is says

Effective search space used: 1344767968614
  Database: NCBI genome chromosomes - human
    Posted date:  Jul 19, 2017  11:08 PM
  Number of letters in database: 64,036,671,579
  Number of sequences in database:  3,505

Perhaps the database also includes some older assemblies and unplaced contigs?

Take a look to see what is included using this command:

blastdbcmd -db human_genomic -entry all -outfmt %i%t

That said NCBI is offering something different on their human genome blast page where I captured the above screenshot from.

Log in to answer this question.