This is a test version of Biostars. For the public version, visit https://www.biostars.org.
RFAM problem. Died at rfam_scan.pl line 281, <GEN1> line 1128517

Hi everyone. I am working with RFAM. It is blocked in some line with three different commands.

perl rfam_scan.pl -blastdb Rfam.fasta Rfam.cm 200pb_Pamicro.fasta --nobig -o rfam_200_nobig.out
Died at rfam_scan.pl line 281, <GEN1> line 1128517.

perl rfam_scan.pl -blastdb Rfam.fasta Rfam.cm 200pb_Pamicro.fasta --bt 0.05 -o rfam_200_0.05.out
Died at rfam_scan.pl line 281, <GEN1> line 1128532.

perl rfam_scan.pl -blastdb Rfam.fasta Rfam.cm 200pb_Pamicro.fasta -o rfam_200.out
Died at rfam_scan.pl line 281, <GEN1> line 1129117.

What will you do in my same situation? Do you find some mistake in this line? I am going to send a email to the rfam-help@ebi.ac.uk too.

See you.

blast ncrnas rfam
• 6,106 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

2 answers

In the latest version of the script (version 1.0, released in 2010), line 281 is just an 'if' test that is checking to see the strandedness of the match. It is possible that the format of Infernal has changed and what is being evaluated isn't numeric. That will cause Perl to die, but with a specific message telling you what happened (at least in recent versions of Perl, I can't conveniently test this with older releases right now).

Unfortunately, the script saves the Infernal results in the tmp directory with just the process id for the name, so debugging is not straightforward. You could change line 260 by replacing the /tmp/$$.res to infernal.$$.out and change line 262 from return /tmp/$$.res to return infernal.$$.out and that should give you a results file in the working directory you can inspect. You'd then want to see if the results match the regex in the Infernal parsing sub of that script. There could also be an issue with your particular results, so posting a couple of lines of the Infernal output might help us see what the issue is.

If the above advice is not clear then you may want to start by letting us know the script and Perl version you are using. Also, please let us know if you solve the issue from the help list so there is no duplication of efforts.

• 0 views • link
updated by Ram Ram • written by SES SES

I am going to try that you have told. Only, you had said that line is 260 instead 280 (so... some newcomer arrives here). I will tell you if some works properly.

My perl version :

This is perl 5, version 18, subversion 2 (v5.18.2) built for x86_64-linux-gnu-thread-multi

Infernal version:

Version 1.1; October 2013
• 0 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

Ok. I changed the part of the script that you told me.

#    system "cmsearch $options $cmfile $fafile > /tmp/$$.res" and die;
    system "cmsearch --tabfile infernal.$$.out $options $cmfile $fafile > /tmp/$$.cmsearch" and die;
#    system "cat /tmp/$$.res";
    return ("infernal.$$.out", "/tmp/$$.cmsearch");
}

and failed again.

perl rfam_scan_copy.pl -blastdb Rfam.fasta Rfam.cm 200pb_Pamicro.fasta --bt 0.05   -o  rfam_200_0.05.out
Died at rfam_scan_copy.pl line 281, <GEN1> line 1128336.

perl rfam_scan_copy.pl -blastdb Rfam.fasta Rfam.cm 200pb_Pamicro.fasta --nobig  -o  rfam_200_nobig.out Died at rfam_scan_copy.pl line 281, <GEN1> line 1128759.

perl rfam_scan_copy.pl -blastdb Rfam.fasta Rfam.cm 200pb_Pamicro.fasta -o  rfam_200.out Died at rfam_scan_copy.pl line 281, <GEN1> line 1128416.

How can I see the infernal regex when you refered to "regex in the Infernal parsing sub of that script"?

Now I can see the /tmp. There are four files (*.seq, *.cm, *cmsearch, *.blast). Is it some of them?

• 0 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

You should see a file called "infernal.N.out" in your working directory, where N is some number. Show a few lines of that file and that will let us know if the format has changed (and if the regex isn't correct). Sorry if that wasn't clear.

• 0 views • link
written by SES SES

Don't worry. You was clear. That is the file doesn't appear everywhere. I'll try the command with sudo.

So. I was reading the *.blast archive... and it is seemed that it is make the blast completely.

At the end, if there are some way to import this. In any case, Thanks in advance.

• 0 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

Note that using "sudo" won't change anything, and be sure you are certain what that does. It is really only needed for installing programs or changing system configurations. I didn't test this, so it's not clear to me why you don't see the file. I think you should just use the latest version though (I didn't know there was one) instead of us modifying an older version.

• 0 views • link
written by SES SES

Please take a look at this recent answer on SeqAnswers, which relates to the same errors. It sounds like you need to update your version of Infernal to get the script working properly, and this would explain the most recent error message you posted about the Infernal commands in the script not being recognized.

Be sure to try the latest version of the script along with the updated version of Infernal when you try this analysis again.

• 1 views • link
updated by Ram Ram • written by SES SES

I am looking that there are several version of rfam_scan.pl_v_1.0.4. I will try some of them, previously with changes.

rfam_scan.pl

Also I have warned you that the infernal software update to 1.1.1 the last week (july 23). Then it is possible That they experienced found some mistakes in the last 1.1 version. That :) I hope it works. :)

http://http://infernal.janelia.org/

• 0 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

Good to know there is a more recent version. Definitely use that, and then modifications can be made if need be. Likely they have fixed the issues though.

• 0 views • link
updated by Ram Ram • written by SES SES

It did not work. The file ./tmp/cmsearch is empty with this information.

Failed to parse command line: No such option "--tabfile".
Usage: cmsearch [options] <cmfile> <seqdb>

where basic options are:
  -h        : show brief help on version and usage
  -g        : configure CM for glocal alignment [default: local]
  -Z <x>    : set search space size in *Mb* to <x> for E-value calculations  (x>0)
  --devhelp : show list of otherwise hidden developer/expert options

To see more help on available options, do cmsearch -h
• 1 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

It appears that the parameters to the program cmsearch, a part of Infernal, are incorrect. I would modify notify the authors about this because it looks like their code is still out of sync with latest Infernal commands (and possibly formats).

• 0 views • link
updated by Ram Ram • written by SES SES

The original script from last version of RFAM v11 (rfam_scan.pl) is wrong. Due to it does not change some options than works in the last infernal software v1.0 but it is not update for infernal v1.1. For example, my rfam is blocked when the cmsearch are working. And /tmp you can see why cmsearch file is stop. The first time was because the option --tabfile instead of --tblout or --ga instead of --cut_ga. Although the last option changed (--cut_ga) work neither. A new error appear Error: Format tag is 'HMMER3/f': unrecognized or not supported.

In sum up, You have to explore the rfam_scan with infernal 1.0. I guess.

• 0 views • link
updated by Ram Ram • written by margxenscienculo margxenscienculo

Thanks a lot for the resource!

• 0 views • link
written by uuzpo • uuzpo

Log in to answer this question.

More posts like this
  • Nothing in rfam_scan rusults.
    written by i19870503 • i19870503
    I use `rfam_scan.pl` to annotate the RNA-seq raw reads. But after run that there is nothing in the results file. The code: ``` perl rfam_scan.pl_v_1.0.4 …