Rfam/infernal help running

[criggs]

Hi all,

I am trying to use rfam to search for sequence homologs to many small RNAs (read lists of about 50,000 sequences).

I have successfully installed infernal and downloaded the rfam database. I am unclear on if/how I need to prepare the rfam database using cmbuild, cmcalibrate, and cmsearch. (I saw this in a tutorial for searching a sequence database with a single covariance model). When I tried this, cmbuild worked but cmcalibrate gave an error. Or, is rfam already in the correct format? If so, what are the steps to run a search? Or can you point me to the appropriate tutorial?

Any help on how to run rfam using infernal would be greatly appreciated!

Thanks very much,
Claire

[bastianwur]

What exactly is the error?

Running it is easy, if you use the perl script provided at the website (rfam_scan.pl).

Command:
perl rfam_scan.pl -blastdb /home/Data/RFAM/Rfam.fasta -o /home/Data/RFAM_output.txt --bt 0.0001 --nobig /home/Data/RFAM/Rfam.cm /home/Data/input.fasta

(evalue: empirical from my site; --nobig: Doesn't scan for the big rRNA subunits; rnammer is way faster for that)

[tokikake]

Hi Claire,

if you have downloaded the Rfam.cm from the Rfam website (ftp://ftp.ebi.ac.uk/pub/databases/Rfam/12.0/), then you only have to run cmpress on the Rfam.cm file to prepare it for cmscan (infernal v1.1).