Mercurial > repos > pjbriggs > weeder2

changeset 2:3c5f10f7dd40 draft

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Updated to tool version 2.0.1 (use data table to locate freqfiles).
author pjbriggs
date Fri, 27 Nov 2015 11:06:28 -0500
parents 571cb77ab9e7
children f19e18ab01b1
files README.rst test-data/weeder2.loc tool-data/weeder2.loc.sample tool_data_table_conf.xml.sample tool_data_table_conf.xml.test weeder2_wrapper.sh weeder2_wrapper.xml
diffstat 7 files changed, 105 insertions(+), 24 deletions(-) [+]
line wrap: on
line diff
--- a/README.rst	Tue Dec 09 11:27:19 2014 -0500
+++ b/README.rst	Fri Nov 27 11:06:28 2015 -0500
@@ -30,7 +30,12 @@
 
     <tool file="weeder2/weeder2_wrapper.xml" />
 
-You will also need to install ``weeder2`` from:
+You also need to make a copy of the ``weeder2.loc`` file (a sample version is
+provided here) which lists the species for which frequency files are available.
+This file should be placed in the ``tool-data`` directory of your Galaxy
+installation.
+
+Additionally you will need to install ``weeder2`` from:
 
 - http://159.149.160.51/modtools/downloads/weeder2.html
 
@@ -41,20 +46,41 @@
 
 Also the directory holding the Weeder2 executables should be on your ``PATH``.
 
+Functional tests
+================
+
 If you want to run the functional tests, copy the sample test files under
 sample test files under Galaxy's ``test-data/`` directory. Then:
 
-    ./run_tests.sh -id trimmomatic
+    ./run_tests.sh -id weeder2
 
 You will need to have set the environment variables above.
 
+Reference Data
+==============
+
+Weeder2 requires reference data in the form of frequency files for each
+species of interest. A set of reference files is provided as part of the
+Weeder2 installation.
+
+Additional frequency files can be generated for novel species using the
+``w2frequency_maker`` utility available via:
+
+- http://159.149.160.51/weederaddons/weeder2freq.html
+
+This page also explains what input data should be used.
+
+The location of the additional frequency files can then be specified by
+adding them to the ``weeder2.loc`` file (see above).
+
 History
 =======
 
 ========== ======================================================================
 Version    Changes
 ---------- ----------------------------------------------------------------------
-2.0.0.0    - Initial version
+2.0.1      - Explicitly specify frequency files in ``weeder2.loc``.
+2.0.0      - Initial version
 ========== ======================================================================
 
 
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/weeder2.loc	Fri Nov 27 11:06:28 2015 -0500
@@ -0,0 +1,5 @@
+HS	Homo sapiens (HS)	.
+MM	Mus musculus (MM)	.
+DM	Drosophila melanogaster (DM)	.
+SC	Saccharomyces cerevisiae (SC)	.
+AT	Arabidopsis thaliana (AT)	.
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool-data/weeder2.loc.sample	Fri Nov 27 11:06:28 2015 -0500
@@ -0,0 +1,35 @@
+# This is a sample file distributed with Galaxy that is used by the
+# weeder2 tool to find the frequency files.
+#
+# The weeder.loc file has this format (white space characters are TAB
+# characters):
+#
+#<species_code>	<Description>	<PathToFreqFilesDir>
+#
+# <PathToFreqFilesDir> should be a period ('.') to indicate the
+# frequency files that come as standard with weeder2; otherwise
+# it should be the full path to the directory holding a custom
+# set of .freq files.
+#
+# For example:
+#
+#AF	Aspergillus fumigatus (AF)	/home/galaxy/weeder/FreqFiles
+#
+# where '/home/galaxy/weeder/FreqFiles' would contain:
+#
+#-rw-rw-r-- 1 pjb pjb 20109718 Oct 23 13:16 AF_ds.10.freq
+#-rw-rw-r-- 1 pjb pjb    47080 Oct 23 13:15 AF_ds.6.freq
+#-rw-rw-r-- 1 pjb pjb  1006374 Oct 23 13:15 AF_ds.8.freq
+#-rw-rw-r-- 1 pjb pjb 19930169 Oct 23 13:16 AF_ss.10.freq
+#-rw-rw-r-- 1 pjb pjb    45262 Oct 23 13:16 AF_ss.6.freq
+#-rw-rw-r-- 1 pjb pjb   984583 Oct 23 13:16 AF_ss.8.freq
+#
+# Entries are already provided for the standard frequency files that
+# are distributed with Weeder2.
+# This file should be placed in galaxy's tool-data directory when the
+# weeder2 tool is installed.
+HS	Homo sapiens (HS)	.
+MM	Mus musculus (MM)	.
+DM	Drosophila melanogaster (DM)	.
+SC	Saccharomyces cerevisiae (SC)	.
+AT	Arabidopsis thaliana (AT)	.
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool_data_table_conf.xml.sample	Fri Nov 27 11:06:28 2015 -0500
@@ -0,0 +1,6 @@
+<tables>
+    <table name="weeder2" comment_char="#">
+        <columns>value, name, path</columns>
+        <file path="tool-data/weeder2.loc" />
+    </table>
+</tables>
\ No newline at end of file
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/tool_data_table_conf.xml.test	Fri Nov 27 11:06:28 2015 -0500
@@ -0,0 +1,6 @@
+<tables>
+    <table name="weeder2" comment_char="#">
+        <columns>value, name, path</columns>
+        <file path="${__HERE__}/test-data/weeder2.loc" />
+    </table>
+</tables>
\ No newline at end of file
--- a/weeder2_wrapper.sh	Tue Dec 09 11:27:19 2014 -0500
+++ b/weeder2_wrapper.sh	Fri Nov 27 11:06:28 2015 -0500
@@ -2,31 +2,40 @@
 #
 # Wrapper script to run weeder2 as a Galaxy tool
 #
-# Usage: weeder_wrapper.sh FASTA_IN SPECIES_CODE MOTIFS_OUT MATRIX_OUT [ ARGS... ]
+# Usage: weeder_wrapper.sh FASTA_IN SPECIES_CODE FREQFILES_DIR MOTIFS_OUT MATRIX_OUT [ ARGS... ]
 #
 # ARGS: one or more arguments to supply directly to weeder2
 #
 # Process command line
 FASTA_IN=$1
 SPECIES_CODE=$2
-MOTIFS_OUT=$3
-MATRIX_OUT=$4
+FREQFILES_DIR=$3
+MOTIFS_OUT=$4
+MATRIX_OUT=$5
 #
 # Other arguments
 ARGS=""
-while [ ! -z "$5" ] ; do
-    ARGS="$ARGS $5"
+while [ ! -z "$6" ] ; do
+    ARGS="$ARGS $6"
     shift
 done
 #
 # Link to input file
 ln -s $FASTA_IN
 #
+# Locate the FreqFiles directory
+if [ $FREQFILES_DIR == "." ] ; then
+    # Use the files in the Weeder2 distribution
+    freqfiles_dir=$WEEDER_FREQFILES_DIR
+else
+    # Alternative location
+    freqfiles_dir=$FREQFILES_DIR
+fi
+#
 # Link to the FreqFiles directory as weeder2 executable
 # expects it to be the same directory
-freqfiles_dir=$WEEDER_FREQFILES_DIR
 if [ -d $freqfiles_dir ] ; then
-    echo "Linking to FreqFiles directory"
+    echo "Linking to FreqFiles directory: $freqfiles_dir"
     ln -s $freqfiles_dir FreqFiles
 else
     echo "ERROR FreqFiles directory not found" >&2
--- a/weeder2_wrapper.xml	Tue Dec 09 11:27:19 2014 -0500
+++ b/weeder2_wrapper.xml	Fri Nov 27 11:06:28 2015 -0500
@@ -1,7 +1,10 @@
-<tool id="motiffinding_weeder2" name="Weeder2" version="2.0.0">
+<tool id="motiffinding_weeder2" name="Weeder2" version="2.0.1">
   <description>Motif discovery in sequences from coregulated genes of a single species</description>
+  <requirements>
+    <requirement type="package" version="2.0">weeder</requirement>
+  </requirements>
   <command interpreter="bash">weeder2_wrapper.sh
-  $sequence_file $species_code
+  $sequence_file $species_code ${species_code.fields.path}
   $output_motifs_file $output_matrix_file
   $strands
   #if $chipseq.use_chipseq
@@ -13,21 +16,12 @@
      -sim $advanced_options.sim_threshold
      -em $advanced_options.em_cycles
   #end if
-</command>
-  <requirements>
-    <requirement type="package" version="2.0">weeder</requirement>
-  </requirements>
+  </command>
   <inputs>
     <param name="sequence_file" type="data" format="fasta" label="Input sequence" />
     <param name="species_code" type="select" label="Species to use for background comparison">
-      <!-- Hard code options for now
-	   See weeder's "organisms.txt" for full list
-      -->
-      <option value="HS">Homo sapiens (HS)</option>
-      <option value="MM">Mus musculus (MM)</option>
-      <option value="DM">Drosophila melanogaster (DM)</option>
-      <option value="SC">Saccharomyces cerevisiae (SC)</option>
-      <option value="AT">Arabidopsis thaliana (AT)</option>
+      <options from_data_table="weeder2">
+      </options>
     </param>
     <param name="strands" label="Use both strands of sequence" type="boolean"
 	   truevalue="" falsevalue="-ss" checked="True"