Entering edit mode
7.0 years ago
m.rhodes
▴
50
I am using makeblastdb to create a database of fungal ITS sequence. When I run the command:
./makeblastdb -in uniteits.fasta -out UNITE_ITS.fasta -dbtype nucl -parse_seqids
It appears to run, fine and I get the following output:
Building a new DB, current time: 05/15/2017 16:18:17
New DB name: /Users/user/Documents/BLAST/ncbi-blast-2.6.0+/bin/UNITE_ITS.fasta
New DB title: uniteits.fasta
Sequence type: Nucleotide
Deleted existing Nucleotide BLAST database named /Users/user/Documents/BLAST/ncbi-blast-2.6.0+/bin/UNITE_ITS.fasta
Keep MBits: T
Maximum file size: 1000000000B
Adding sequences from FASTA; added 29342 sequences in 1.83726 seconds.
However, when I search for the databases, it is not there. I double checked that it wasn't just somewhere else on my computer, but still nothing. I've read some previous posts suggesting that the issue may be something to do with permissions, but did not provide a solid answer. I tried using chmod, if this was in fact the case, but to no avail.
Can anyone help me please?
Thanks
Your database file should be here:
/Users/user/Documents/BLAST/ncbi-blast-2.6.0+/bin/UNITE_ITS.fasta
This also isn't the first time you've run this command with this output name, because the previous version was replaced:
Deleted existing Nucleotide BLAST database named /Users/user/Documents/BLAST/ncbi-blast-2.6.0+/bin/UNITE_ITS.fasta
Is
uniteits.fasta
in the bin folder for blast?I know, I've tried running the command several times after playing around with changing permission, but each time there hasn't been any output in the bin folder or anywhere else on my laptop. Yes, the uniteits.fasta is in bin folder for blast.
double check that your input file is in fasta format. You can use
-input_type fasta
to specify, and also if you have long enough headers to parse.The programme won't run if its not a fasta (I've accidentally done it in non-fasta format before and it gave me an error), so I'm pretty sure that isn't the issue. As for the headers, I followed the specification for makeblastdb header that was on the NCBI site so I doubt that is the problem either.
What is the outout of
ls -lh /Users/user/Documents/BLAST/ncbi-blast-2.6.0+/bin/UNITE_ITS.fasta*
?So the database has definitely been created. How many input sequences were there in your fasta?
What blast+ command did you use for the search?
There should be 29,342 sequences in total. The command I used was:
./makeblastdb -in uniteits.fasta -out UNITE_ITS.fasta -dbtype nucl -parse_seqids
You can check the integrity/validity of the blast database with
blastdbcheck
program. Seeblastdbcheck -h
for command options.I was asking if you had tried to use this database for an actual blast search (and the command you used)?
No I hadn't tried to use it for a blast search yet as I think that any of any of the outputs I got were correct. I'm still unsure on the difference between the 6 outputs.
That is what a set of blast index files look like. Make sure you put a copy of the original fasta file (or create a symlink) in the directory where these files are.
Ok, the suggestion there however is that I use the UNITE_ITS.fasta file as the database to blast against. This file does not exist.
I am always puzzled as to why people choose to use
something.fasta
as the name for their blast database ..In any case keep a copy (or symlink)
uniteits.fasta
in the blast index files directory. Try a real blast run and let us know how it goes.