Re: TCoffee Newsgroup: Issues with legacy_blast.pl when using psicoffee mode

[Cedric Notredame]

This looks about right to me. One question : is your database formatted for Blast. This needs to be done before hand  using formatdb (I think)

-- 

Dr Cedric Notredame, PhD
ORCID - https://orcid.org/0000-0003-1461-0988

On 19 Sep 2024, at 19:08, fer...@gmail.com <fer...@gmail.com> wrote:

Hello group,

I have a transmembrane protein dataset, for which I want to use psicoffee for alignment. After browsing and digging around, I came up with the command:

t_coffee Pl-B2_s5_cdh70.fsa -mode psicoffee -blast_server LOCAL -protein_db /home/fer/Desktop/test/uniref90_TM/uniref90_TM.fasta

but I get:

INPUT FILES
Input File (M) proba_pair
Input File (S) Pl-B2_s5_cdh70.fsa  Format fasta_seq

Identify Master Sequences [no]:

Master Sequences Identified
Looking For Sequence Templates:

Template Type: [BLAST] Mode Or File: [BLAST] [Start
#*****************************************************************
--ERROR: #  [FATAL:T-COFFEE]

# The Program legacy_blast.pl Needed by T-COFFEE Could not be found
# If legacy_blast.pl is installed on your system:
#      -Make sure legacy_blast.pl is in your $path:
# If legacy_blast.pl is NOT installed obtain a copy from:
# ftp://ftp.ncbi.nih.gov/blast/executables/LATEST
#
#
# and install it manualy
******************************************************************


*************************************************************************************************
*                        FULL TRACE BACK PID: 46977                                    
46977 -- ERROR: #  [FATAL:T-COFFEE]
46977 -- COM: t_coffee Pl-B2_s5_cdh70.fsa -mode psicoffee -blast_server LOCAL -protein_db /home/fer/Desktop/test/uniref90_TM/uniref90_TM.fasta
46977 -- STACK: 46930 -> 46977
*************************************************************************************************

# TERMINATION STATUS: FAILURE [PROGRAM: T-COFFEE pid 46977 ppid 46930
#CL: t_coffee Pl-B2_s5_cdh70.fsa -mode psicoffee -blast_server LOCAL -protein_db /home/fer/Desktop/test/uniref90_TM/uniref90_TM.fasta

I have already manually installed the package containing legacy_blast.pl, which I have also added to $PATH (I checked and it's dicoverable). Also, the database was formatted with makeblastdb.

So, any ideas on what may be going on here? Thanks in advance! 

--
You received this message because you are subscribed to the Google Groups "Tcoffee" group.
To unsubscribe from this group and stop receiving emails from it, send an email to tcoffee+u...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/tcoffee/6d816316-f89c-4c92-9106-5048fb5e57f3n%40googlegroups.com.

[fer...@gmail.com]

Thanks Cedric for the reply. Yes indeed, the db is formatted, although I've used makeblastdb to do so, since I understand it's a "newer version" of formatdb, and I've used it many times before successfully. I will try out formatdb though.

Still the critical error seems to be linked to legacy_blast.pl, which as I mentioned before I've downloaded, installed and placed in $PATH. Maybe is something about how t_coffee calls BLAST? Is there a way I could test that?

Thanks! Fer

[fer...@gmail.com]

Final update: it's running now. 

In case it helps somebody else: It wasn't needed to reformat the database (makeblastdb worked fine). But I already had Blast+ intalled in my system, so I deleted the package completely via apt-get. Then, I modified the $PATH to the folder containing legacy_blast.pl and all the blast executables manually installed from the package downloaded from NCBI's ftp (which in my case is the folder /home/fer/ncbi-blast-2.16.0+/bin/). 

So now this is the only BLAST package I have in my system, and psicoffee is running. 

Thanks! Fer

[Cedric Notredame]

Awesome. Thanks Fer! We should probably check the blast version and return a more meaningful error msg. 

-- 

Dr Cedric Notredame, PhD
ORCID - https://orcid.org/0000-0003-1461-0988

On 20 Sep 2024, at 15:18, fer...@gmail.com <fer...@gmail.com> wrote:

Final update: it's running now. 

To view this discussion on the web visit https://groups.google.com/d/msgid/tcoffee/5579048e-aff6-4cb8-8fa3-021ad7b041f7n%40googlegroups.com.