PfamScanner

PfamScanner is a wrapper around hmmscan for scanning against protein sequences against Pfam-defined domain models. Scan output is provided in a fasta-like format (xdom).

The .xdom format

For example, the Human ATP-binding cassette sub-family F member 2 (Q9UG63) contains three Pfam-A domains:

>Q9UG63  623
44  257  ABC_tran
296 383  ABC_tran_2
437 545  ABC_tran	

The first line specifies the protein ID and its length in amino acids. The remaining lines specify each domain (in sequence) with its repective co-ordinates and ID or accession number. Optionally, the ID fields can be followed by an Evalue inidcating the significance of the hit between the domain sequence and the defining model.

Features

• Runs in scan / parse-only mode
• Provides overlap resolution (by best matching cascade)
• Can collapse successive domains of type repeat (repeat arrays)
• Can merge split hits (successive hits, in sequence, to the same model)

Usage

$ java -jar build/PfamScanner.jar

Usage: PfamScanner [OPTIONS] -in <infile> -out <outfile> -dir <workingdir>
Run hmmscan against Pfam defined domains

 -acc,--accession       Use Pfam (PF00002) accessions instead of IDs
                        (7tm_2)
 -c,--cpu <arg>         Number of parallel CPU workers to use for
                        multithreads (hmmscan)
 -C,--collapse          Collapse domains of type repeat
 -dir <directory>       The working directory. This directory must contain
                        the Pfam-A domain models, pressed using hmmpress
                        and named Pfam-A.hmm. This directory must also
                        contain the hmmscan binary (version >= 3.0). By
                        default, the working directory is set to the
                        current directory.
 -e,--evalue <float>    Evalue threshold [Default: model defined gathering
                        threshold]
 -h,--help              Print this help message
 -in <file>             Fasta || hmmout input file (see option parse-only)
 -M,--model <models>    File containing HMMs for scan (must be indexed)
 -m,--merge             Merge split hits
 -out <file>            XDOM output file
 -p,--parse-only        Parse previous run of hmmscan (save to file). If
                        set, <in> file must be hmmscan (version 3)
                        domtblout format, and <out> the name of the xdom
                        that should be written to.
 -r,--remove-overlaps   Resolve overlaps (Best match cascade)
 -s,--save <file>       Save hmmscan output (full path required)
 -t,--tempdir <dir>     Directory in which temporary files are to be
                        written.
 -v,--verbose           Verbose scan