Mercurial > repos > dawe > srf2fastq
view srf2fastq/README @ 0:d901c9f41a6a default tip
Migrated tool version 1.0.1 from old tool shed archive to new tool shed repository
author | dawe |
---|---|
date | Tue, 07 Jun 2011 17:48:05 -0400 |
parents | |
children |
line wrap: on
line source
This tool comes with io_lib (from Staden package) version 1.12.2 io_lib is necessary as you will need srf2fastq executable; if you alread have the binary, skip the io_lib installation. You may find a more recent version here http://sourceforge.net/projects/staden/files/ 1- Install io_lib cd io_lib-1.12.2 ./configure --prefix=PREFIX make make install #assuming you have already superuser permissions granted 2- Edit srf2fastq.sh You will need to change the path of srf2fastq executable if PREFIX has been defined different from /usr/local:wq 3- Add support for SRF files in galaxy (if galaxy team hasn't added yet) Here comes a series of diffs you may want to integrate. lib/galaxy/datatypes/binary.py and lib/galaxy/datatypes/registry.py are part of galaxy distribution so a hg diff output is given. datatype_conf.xml is not in galaxy repository (the sample is), so a unified diff is given (well, it's the same...) diff -r 60448575467f lib/galaxy/datatypes/binary.py --- a/lib/galaxy/datatypes/binary.py Mon Aug 23 10:15:17 2010 -0400 +++ b/lib/galaxy/datatypes/binary.py Wed Aug 25 10:26:38 2010 +0200 @@ -222,3 +222,31 @@ Binary.__init__( self, **kwd ) self._magic = 0x8789F2EB self._name = "BigBed" + +class Srf( Binary ): + """ Illumina Sequence Archive (SRF) """ + file_ext = "srf" + + def __init__( self, **kwd ): + Binary.__init__( self, **kwd ) + def sniff( self, filename ): + try: + header = open( filename ).read(4) + if binascii.b2a_hex( header ) == binascii.hexlify( 'SSRF' ): + return True + return False + except: + return False + def set_peek( self, dataset, is_multi_byte=False ): + if not dataset.dataset.purged: + dataset.peek = "Binary srf file" + dataset.blurb = data.nice_size( dataset.get_size() ) + else: + dataset.peek = 'file does not exist' + dataset.blurb = 'file purged from disk' + def display_peek( self, dataset ): + try: + return dataset.peek + except: + return "Binary srf file (%s)" % ( data.nice_size( dataset.get_size() ) ) + diff -r 60448575467f lib/galaxy/datatypes/registry.py --- a/lib/galaxy/datatypes/registry.py Mon Aug 23 10:15:17 2010 -0400 +++ b/lib/galaxy/datatypes/registry.py Wed Aug 25 10:28:01 2010 +0200 @@ -173,6 +173,7 @@ 'sam' : tabular.Sam(), 'scf' : binary.Scf(), 'sff' : binary.Sff(), + 'srf' : binary.Srf(), 'tabular' : tabular.Tabular(), 'taxonomy' : tabular.Taxonomy(), 'txt' : data.Text(), @@ -213,6 +214,7 @@ if len(self.sniff_order) < 1: self.sniff_order = [ binary.Bam(), + binary.Srf(), binary.Sff(), xml.BlastXml(), sequence.Maf(), diff -u datatypes_conf.xml datatypes_conf.xml.sample --- datatypes_conf.xml 2010-08-25 10:32:46.778959931 +0200 +++ datatypes_conf.xml.sample 2010-08-25 07:20:34.190956591 +0200 @@ -105,7 +105,6 @@ <datatype extension="scf" type="galaxy.datatypes.binary:Scf" mimetype="application/octet-stream" display_in_upload="true"/> <datatype extension="Sequences" type="galaxy.datatypes.assembly:Sequences" display_in_upload="false"/> <datatype extension="sff" type="galaxy.datatypes.binary:Sff" mimetype="application/octet-stream" display_in_upload="true"/> - <datatype extension="srf" type="galaxy.datatypes.binary:Srf" mimetype="application/octet-stream" display_in_upload="true"/> <datatype extension="taxonomy" type="galaxy.datatypes.tabular:Taxonomy" display_in_upload="true"/> <datatype extension="tabular" type="galaxy.datatypes.tabular:Tabular" display_in_upload="true"/> <datatype extension="txt" type="galaxy.datatypes.data:Text" display_in_upload="true"/> @@ -264,7 +263,6 @@ and so on. --> <sniffer type="galaxy.datatypes.binary:Bam"/> - <sniffer type="galaxy.datatypes.binary:Srf"/> <sniffer type="galaxy.datatypes.binary:Sff"/> <sniffer type="galaxy.datatypes.xml:BlastXml"/> <sniffer type="galaxy.datatypes.sequence:Maf"/>