Domain annotation

To annotate sequences with domains, different programs and databases exist. Our programs mainly support Pfam domains. This section will give you a short overview on how to annoate your protein sequences with pfam domains. The standard program to annotate your proteome with pfam domains is pfam_scan.pl which is provided by the Pfam database.

Requirements

Preparing the database

This steps needs to be done only once when you downloaded a new version of the HMM Models from Pfam.

Downloaded the Pfam-A.hmm.gz and the Pfam-A.hmm.dat.gz file from the Pfam release you want to use: http://ftp.ebi.ac.uk/pub/databases/Pfam/releases/. Then, you will need to run the following commands to create a database that is usable by pfam_scan.pl/hmmscan

# unzip the downloaded files
gunzip Pfam-A.hmm.gz
gunzip Pfam-A.hmm.dat.gz

# use hmmpress from the HMMER suite to prepare the databse for hmmscan
hmmpress Pfam-A.hmm

These steps will create some additional files.

Running pfam_scan.pl

For pfam_scan.pl to work properly you will need to:

  1. Have the HMMER binaries in your PATH variable for example by manipulating PATH

export PATH=<path to directory containing HMMER executables>:$PATH

  1. Set the PERL5LIB variable to include the pfam_scan.pl directory

export PERL5LIB=<path to PfamScan directory>:$PERL5LIB

You can then call pfam_scan.pl

pfam_scan.pl -fasta <sequence file> -dir <directory containing the hmm files> -outfile <output file>