# HG changeset patch
# User bgruening
# Date 1428662862 14400
# Node ID 53c20f28562edc2896112e3604a4ea0e9dd9d848
# Parent  97ecd7154f6b6c0ed956bceb7098a40d5ed8de81
Uploaded

diff -r 97ecd7154f6b -r 53c20f28562e .shed.yml
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/.shed.yml	Fri Apr 10 06:47:42 2015 -0400
@@ -0,0 +1,3 @@
+# repository published to https://toolshed.g2.bx.psu.edu/repos/bgruening/find_subsequences
+owner: bgruening
+name: find_subsequences
diff -r 97ecd7154f6b -r 53c20f28562e find_subsequences.py
--- a/find_subsequences.py	Fri Mar 20 06:21:30 2015 -0400
+++ b/find_subsequences.py	Fri Apr 10 06:47:42 2015 -0400
@@ -10,7 +10,7 @@
 
 choices = ['embl', 'fasta', 'fastq-sanger', 'fastq', 'fastq-solexa', 'fastq-illumina', 'genbank', 'gb']
 
-def find_pattern(seqs, pattern, outfile_path):
+def find_pattern(seqs, pattern, outfile_path, strand):
     """
     Finds all occurrences of a pattern in the a given sequence.
     Outputs sequence ID, start and end postion of the pattern.
@@ -23,8 +23,10 @@
 
     with open(outfile_path, 'w+') as outfile:
         for seq in seqs:
-            search_func(seq, pattern, outfile)
-            search_func(seq, rev_compl, outfile, '-')
+            if strand in ['both', 'forward']:
+                search_func(seq, pattern, outfile)
+            if strand in ['both', 'reverse']:
+                search_func(seq, rev_compl, outfile, '-')
 
 
 def simple_pattern_search(sequence, pattern, outfile, strand='+'):
@@ -50,12 +52,13 @@
 
 if __name__ == "__main__":
     parser = argparse.ArgumentParser()
-    parser.add_argument('-i', '--input' , required=True)
-    parser.add_argument('-o', '--output' , required=True)
-    parser.add_argument('-p', '--pattern' , required=True)
+    parser.add_argument('-i', '--input', required=True)
+    parser.add_argument('-o', '--output', required=True)
+    parser.add_argument('-p', '--pattern', required=True)
+    parser.add_argument('--strand', choices=['both', 'forward', 'reverse'], default='both')
     parser.add_argument('-f', '--format', default="fasta", choices=choices)
     args = parser.parse_args()
 
     with open(args.input) as handle:
-        find_pattern( SeqIO.parse(handle, args.format), args.pattern, args.output )
+        find_pattern( SeqIO.parse(handle, args.format), args.pattern, args.output, args.strand )
 
diff -r 97ecd7154f6b -r 53c20f28562e find_subsequences.xml
--- a/find_subsequences.xml	Fri Mar 20 06:21:30 2015 -0400
+++ b/find_subsequences.xml	Fri Apr 10 06:47:42 2015 -0400
@@ -6,14 +6,15 @@
     <command interpreter="python">
     <![CDATA[
         find_subsequences.py
-            --input $input
-            --output $output
+            --input "${input}"
+            --output "${output}"
             --pattern "$pattern_conditional.pattern"
             #if $input.ext == 'fasta':
                 --format 'fasta'
             #else:
                 --format 'fastq'
             #end if
+            --strand '$strand'
     ]]>
     </command>
     <inputs>
@@ -248,6 +249,11 @@
                 </param>
             </when>
         </conditional>
+        <param name="strand" type="select" label="Search pattern on">
+            <option value='both'>both strands</option>
+            <option value='forward'>forward strand</option>
+            <option value='reverse'>reverse strand</option>
+        </param>
     </inputs>
     <outputs>
         <data format="bed" name="output" />
@@ -271,6 +277,20 @@
             <param name="pattern_conditional_select" value="user"/>
             <output name="output" file="find_subsequences_advanced_result1.bed" ftype="bed"/>
         </test>
+        <test>
+            <param name="input" value="find_subsequences_input1.fasta" ftype="fasta"/>
+            <param name="pattern" value="atnncg"/>
+            <param name="strand" value="reverse"/>
+            <param name="pattern_conditional_select" value="user"/>
+            <output name="output" file="find_subsequences_advanced_result2.bed" ftype="bed"/>
+        </test>
+        <test>
+            <param name="input" value="find_subsequences_input1.fasta" ftype="fasta"/>
+            <param name="pattern" value="atnncg"/>
+            <param name="strand" value="forward"/>
+            <param name="pattern_conditional_select" value="user"/>
+            <output name="output" file="find_subsequences_advanced_result3.bed" ftype="bed"/>
+        </test>
     </tests>
     <help>
 <![CDATA[
diff -r 97ecd7154f6b -r 53c20f28562e test-data/find_subsequences_advanced_result2.bed
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/find_subsequences_advanced_result2.bed	Fri Apr 10 06:47:42 2015 -0400
@@ -0,0 +1,1 @@
+reverse_advanced	9	15	reverse_advanced		-
diff -r 97ecd7154f6b -r 53c20f28562e test-data/find_subsequences_advanced_result3.bed
--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/test-data/find_subsequences_advanced_result3.bed	Fri Apr 10 06:47:42 2015 -0400
@@ -0,0 +1,1 @@
+forward_advanced	9	15	forward_advanced		+