[Inparanoid] BLASTALL and inparanoid

[Kristoffer Forslund]

Most of the time for large applications (such as when building the
database), we would indeed run blast separately the way you suggest. The
only thing is to check for:

- The 2-pass blast option requires some extra scripting to implement
separately, if you want to do that.
- You should make sure that the various flags do the same things in the
blast version you want to use and:
- You should check that the XML output still is correctly handled by the
parsing script.

My guess is that the blastp application should be more or less exactly the
same as the blastall one in terms of input and output, in which case you
could just change the lines in the inparanoid.pl script where it is
invoked to use the newer executable instead. As long as the flags passed
to it still do the same thing (which you could check using the blast
syntax description), there should be no problem that I can see.

Yours sincerely,

Kristoffer Forslund

> Hello Tom,
>
> I am not affiliated with InParanoid, and I am only an end user like
> yourself. However, I have a suggestion that might prove helpful, if you
> are willing to tinker a bit. (It worked for me, but I'd be curious if the
> InParanoid developers advise against this for some reason...)
>
> I ran BLAST independently of InParanoid, as I had many proteomes to
> compare and I wanted to distribute the BLAST jobs on my institute's grid.
> I then used the BLAST outputs as inputs into InParanoid's blast_parser.pl
> script, and finally ran InParanoid. To achieve this, I did the following:
>
> 1. Run BLAST by whatever means you like, specifying to generate outputs in
> XML format. (I used that blastall program, but I don't see why this won't
> work with blast+ provided you specify XML output. For blastall, I
> specified these parameters, in addition to the query and database: -e 1e-1
> -F 'm S' -b 20 -v 20 -M BLOSUM62 -m 7)
>
> 2. Feed the resulting BLAST XML files into the blast_parser.pl script that
> comes with InParanoid, as follows. Notes: (a) my FASTA files and BLAST
> outputs are named without extensions; for example, species a is called "a"
> (not "a.fa") and species a vs. species b is simply called "ab" (not
> "ab.xml"). (b) All BLAST outputs are in a sub-directory called
> "blast_results". (c) To be safe, I backed up my BLAST results before I
> tried this, since it takes a long time to run the BLAST searches...
>
> In a BASH terminal (command line), from the main directory, do this:
>
> cat blast_results/ab | ./blast_parser.pl 0 > ab
>
> This will read the BLAST output called "ab" from the sub directory
> "blast_results" and pass it to the blast_parser.pl script in the main
> directory, which will generate an output file called "ab" in the main
> directory. I parsed multiple XML files with a simple BASH script run from
> the main directory to automate the process, like this:
>
>  #!/bin/bash
>  for i in aa ab bb ba ac bc # [et cetera - add any number of files here]
>  do
>  cat blast_results/$i | ./blast_parser.pl 0 > $i
>  done
>
> 3. Set the variable to turn off BLAST inside the InParanoid file (line 44
> or 45): $run_blast = 0
>
> 4. Run InParanoid (using outgroup in my case): ./inparanoid.pl a b c
>
> I hope that is helpful.
>
> Best regards,
> Marcus Chibucos
>
> On 5/24/12 10:15 AM, de Man, Tom wrote:
> Dear people from inparanoid,
>
> The first step of inparanoid is performing a BLASTALL run, however the new
> BLAST+ does not contain BLASTALL anymore.
> With this reimplementation of BLAST you deal with separate programs.
>
> At the original BLAST you use commands like: BLASTALL -p blastp. However
> with BLAST+ you can run blastp as a program instead of using it as a
> parameter of BLASTALL. Is there a version of inparanoid that can handle
> this new BLAST or should I just use an older version of BLAST with
> BLASTALL included.
>
> Many thanks for your reply.
>
> Kind regards,
> Tom de Man
>
> ________________________________
>
> This e-mail message (including any attachments) is for the sole use of
> the intended recipient(s) and may contain confidential and privileged
> information. If the reader of this message is not the intended
> recipient, you are hereby notified that any dissemination, distribution
> or copying of this message (including any attachments) is strictly
> prohibited.
>
> If you have received this message in error, please contact
> the sender by reply e-mail message and destroy all copies of the
> original message (including attachments).
> _______________________________________________
> InParanoid mailing list
> InParanoid at sbc.su.se
> https://mail.sbc.su.se/mailman/listinfo/inparanoid
>