From d0c13d7ec425cefd76aba3e7d6373ae03ecc5ef5 Mon Sep 17 00:00:00 2001
From: Guruprasad Anada <gua110@bx.psu.edu>
Date: Fri, 3 Apr 2009 13:25:01 -0400
Subject: [PATCH] Added a new option to specify minimum match value in the
 liftOver tool.

---
 tools/extract/liftOver_wrapper.py  | 12 ++++++++----
 tools/extract/liftOver_wrapper.xml | 10 ++++++----
 2 files changed, 14 insertions(+), 8 deletions(-)

diff --git a/tools/extract/liftOver_wrapper.py b/tools/extract/liftOver_wrapper.py
index a9f88d4af05..f15fead0d2b 100644
--- a/tools/extract/liftOver_wrapper.py
+++ b/tools/extract/liftOver_wrapper.py
@@ -12,20 +12,24 @@ def stop_err(msg):
     sys.stderr.write(msg)
     sys.exit()
     
-if len( sys.argv ) != 6:
-    stop_err( "USAGE: prog input out_file1 out_file2 input_dbkey output_dbkey" )
+if len( sys.argv ) != 7:
+    stop_err( "USAGE: prog input out_file1 out_file2 input_dbkey output_dbkey minMatch" )
 
 infile = sys.argv[1]
 outfile1 = sys.argv[2]
 outfile2 = sys.argv[3]
 in_dbkey = sys.argv[4]
 mapfilepath = sys.argv[5]
-
+minMatch = sys.argv[6]
+try:
+    assert float(minMatch)
+except:
+    minMatch = 0.1
 #ensure dbkey is set
 if in_dbkey == "?": 
     stop_err( "Input dataset genome build unspecified, click the pencil icon in the history item to specify it." )
 
-cmd_line = "liftOver " + infile + " " + mapfilepath + " " + outfile1 + " " + outfile2 + "  > /dev/null 2>&1"
+cmd_line = "liftOver -minMatch=" + minMatch + " " + infile + " " + mapfilepath + " " + outfile1 + " " + outfile2 + "  > /dev/null 2>&1"
 
 if not os.path.isfile( mapfilepath ):
     stop_err( "%s mapping is not currently available."  % ( mapfilepath.split('/')[-1].split('.')[0] ) )
diff --git a/tools/extract/liftOver_wrapper.xml b/tools/extract/liftOver_wrapper.xml
index 99bc7dee46b..a19784a2a21 100644
--- a/tools/extract/liftOver_wrapper.xml
+++ b/tools/extract/liftOver_wrapper.xml
@@ -1,11 +1,11 @@
 <tool id="liftOver1" name="Convert genome coordinates">
   <description> between assemblies and genomes</description>
-  <command interpreter="python">liftOver_wrapper.py $input "$out_file1" "$out_file2" $dbkey $to_dbkey</command>
+  <command interpreter="python">liftOver_wrapper.py $input "$out_file1" "$out_file2" $dbkey $to_dbkey $minMatch</command>
   <inputs>
     <param format="interval" name="input" type="data" label="Convert coordinates of">
       <validator type="unspecified_build" />
       <validator type="dataset_metadata_in_file" filename="liftOver.loc" metadata_name="dbkey" metadata_column="0" message="Liftover mappings are currently not available for the specified build." />
-    </param>
+    </param>
     <param name="to_dbkey" type="select" label="To">
       <options from_file="liftOver.loc">
         <column name="name" index="1"/>
@@ -13,7 +13,8 @@
         <column name="dbkey" index="0"/>
         <filter type="data_meta" ref="input" key="dbkey" column="0" />
       </options>
-    </param> 
+    </param> 
+    <param name="minMatch" size="10" type="float" value="0.10" label="Minimum ratio of bases that must remap" />
   </inputs>
   <outputs>
     <data format="input" name="out_file1" />
@@ -26,7 +27,8 @@
   <tests>
     <test>
       <param name="input" value="5.bed" dbkey="hg18" ftype="bed" />
-      <param name="to_dbkey" value="/depot/data2/galaxy/hg18/liftOver/hg18ToPanTro2.over.chain" />
+      <param name="to_dbkey" value="/depot/data2/galaxy/hg18/liftOver/hg18ToPanTro2.over.chain" />
+      <param name="misMatch" value="0.95" />
       <output name="out_file1" file="5_liftover_mapped.bed"/>
       <output name="out_file2" file="5_liftover_unmapped.bed"/>
     </test>