3
4
Entering edit mode
10.5 years ago
diltsjeri ▴ 470

What is the best way to download GenBank locally? What is the best method for creating databases with the data? BioPerl? Thanks!

genbank local • 18k views
8
Entering edit mode
10.5 years ago
Lee Katz ★ 3.1k

See ftp://ftp.ncbi.nlm.nih.gov/blast/documents/blastdb.html for details.

Part two of your question: usually the best kind of database is a blast database.

2
Entering edit mode

Perfect.Thanks!

As a side note. The link up there is not working, but this should take you to where you need to go http://www.ncbi.nlm.nih.gov/BLAST/docs/update_blastdb.pl (for anyone looking at this thread)

0
Entering edit mode

Sweet i need this too, this tar has the latest scripts and tools to use blast locally ftp://ftp.ncbi.nlm.nih.gov/blast/executables/blast+/LATEST/

0
Entering edit mode
10.5 years ago
jrymer1 ▴ 30

I'm getting an error when running the updateblastdb.pl. I do perl update_blastdb.pl and get the error that perl docs needs to be installed, so installed it but now when I do the same command I'm getting a different page giving me options. So no I'm using perl update_blastdb.pl -showall blastdb and I get the error "use of uninitialized value $retval[0] in concatenation (.) or string at updateblast.db line 110. How do I run this script to download the BLAST database? ADD COMMENT 1 Entering edit mode Also when I do 'perl update_blastdb.pl blast db' I get the error "blastdb not found skipping." Is there an alternate way to download the entire BLAST database other than this? ADD REPLY 2 Entering edit mode You need to know the name of the database you want to download. For example, run: perl update_blastdb.pl nr  Also, it's not really proper to use the "answer" section to ask your own questions. ADD REPLY 0 Entering edit mode I tried with the est_mouse database. I did perl update_blastdb.pl est_mouse. and got the error est_mouse not found, skipping. The same happened with nr, I tried your exact line. Im running on Ubuntu 12.04 in virtual box, would that be the cause of the issue? ADD REPLY 1 Entering edit mode That's strange. I just started the est_mouse command and it seems to be running fine. It also works okay on my 12.04 Ubuntu bare install (just tested), but it seems like it is failing in the virtual image I just spun up (vanilla install of 12.04). I am not entirely sure what is going on. If all else fails, you can download it from their FTP site directly: ftp::ftp.ncbi.nlm.nih.gov/blast/db ADD REPLY 0 Entering edit mode I got the following error: Connected to NCBI Downloading nr (46 volumes) ... Downloading nr.00.tar.gz...Cannot open Local file nr.00.tar.gz: Permission denied at update_blastdb.pl line 202 Cannot open Local file nr.00.tar.gz.md5: Permission denied at update_blastdb.pl line 203 Failed to download nr.00.tar.gz.md5!  Would you mind please let me know what's the problem? ADD REPLY 0 Entering edit mode Hi. Can you check that file (nr.00.tar.gz.md5) and see what is with it? Is its size 0? Has it the correct permissions? ADD REPLY 0 Entering edit mode Although it has been a few years. I ran into the same issue! I am not used to windows and installed blast under C:\Programs. What I didn't know is, that under windows you will never be able to set write-rights in this directory on a specific script. Just run the script in a different directory other than C:\Programs. mkdir C:\blastdbs cd C:\blastdbs\ update_blastdb.pl --decompress nt  ADD REPLY 1 Entering edit mode It has been roughly 6.3 years since this was posted, but in case you or anyone is still wondering, I think your issue is the use of the perl command at the beginning. with the installation of BLAST+ the path to update_blastdb.pl is likely added to the PATH variable. The perl command takes a script in the current directory and runs it, however, the update_blastdb.pl document can be ran without the use of perl (as it initiates perl by itself already). therefore you can run the following update_blastdb.pl --help  or update_blastdb.pl <dbname> update_blastdb.pl nt update_blastdb.pl taxdb  This will download the *.tar.gz files from the NCBI website. However, you still have to decompress them with either the --decompress option of update_blastdb.pl or with tar: update_blastdb.pl nt taxdb --decompress for subdb in *.tar.gz; do tar -zxvf$subdb; done


With enough processors one might consider using an "&" though, as this runs the extraction parallel for all files

for subdb in *.tar.gz; do tar -zxvf \$subdb & done


------EDIT------

It is worthwhile realizing that the blastn command must be ran from within the directory where the database is saved. If only the path is specified blastn will not be able to output taxonomic / scientific names.

0
Entering edit mode
7.9 years ago
BioApps ▴ 790

Here is a tool for Windows and Mac users that allows you to download some or all databases: