changeset 3:30b17a17c1aa draft

Uploaded

--- /dev/null	Thu Jan 01 00:00:00 1970 +0000
+++ b/call_amr_mutations.py	Tue Mar 21 19:26:55 2023 +0000
@@ -0,0 +1,37 @@
+#!/usr/bin/env python
+
+# NOTE: This tool provides the functionality of the PIMA filter_varsacn()
+# function here https://github.com/appliedbinf/pima_md/blob/main/pima.py#L3012
+
+import argparse
+import re
+import subprocess
+import sys
+
+
+def run_command(self, command):
+    try:
+        return re.split('\\n', subprocess.check_output(command, shell=True).decode('utf-8'))
+    except Exception:
+        message = 'Command %s failed: exiting...' % command
+        sys.exit(message)
+
+
+def filter_varscan(varscan_raw, output):
+    cmd = ' '.join(['cat', varscan_raw,
+                    '| awk \'(NR > 1 && $9 == 2 && $5 + $6 >= 15)',
+                    '{OFS = "\\t";f = $6 / ($5 + $6); gsub(/.*\\//, "", $4);s = $4;gsub(/[+\\-]/, "", s);$7 = sprintf("%.2f%%", f * 100);'
+                    'min = 1 / log(length(s) + 2) / log(10) + 2/10;if(f > min){print}}\'',
+                    '1>' + output])
+    output = run_command(cmd)
+
+
+if __name__ == '__main__':
+    parser = argparse.ArgumentParser()
+
+    parser.add_argument('--varscan_raw', action='store', dest='varscan_raw', help='Raw varscan mpileup VCF file')
+    parser.add_argument('--output', action='store', dest='output', help='Output filtered VCF file')
+
+    args = parser.parse_args()
+
+    filter_varscan(args.varscan_raw, args.output)

--- a/call_amr_mutations.xml	Mon Feb 20 21:10:42 2023 +0000
+++ b/call_amr_mutations.xml	Tue Mar 21 19:26:55 2023 +0000
@@ -5,33 +5,29 @@
     </macros>
     <expand macro="requirements"/>
     <command detect_errors="exit_code"><![CDATA[
-#import os
+## NOTE: This tool provides the functionality of the PIMA filter_varsacn()
+## function here https://github.com/appliedbinf/pima_md/blob/main/pima.py#L3012
 
-#if $os.path.getsize(str($varscan_raw)) > 0:
-    cat '$varscan_raw' | awk '(NR > 1 && $9 == 2 && $5 + $6 >= 15){OFS = "\t";f = $6 / ($5 + $6); gsub(/.*\//, "", $4);s = $4;gsub(/[+\\-]/, "", s);$7 = sprintf("%.2f%%", f * 100);min = 1 / log(length(s) + 2) / log(10) + 2/10;if(f > min){print}}' > varscan_snp &&
-    cat varscan_snp | awk '{OFS = "\t"; print $1,$2,".",$3,$4,-log($14),"PASS",".","GT","1|1"}' > snp_vcf &&
-    cat snp_vcf | sort -k 1,1 -k 2n,2n | awk 'BEGIN{OFS = "\t";print "#CHROM\tPOS\tID\tREF\tALT\tQUAL\tFILTER\tINFO\tFORMAT\tSAMPLE"}{print}' > '$varscan_vcf'
-#end if
+python '$__tool_directory__/filter_varscan.py'
+--varscan_raw '$varscan_raw'
+--output '$output'
 ]]></command>
     <inputs>
         <param name="varscan_raw" type="data" format="vcf" label="Raw varscan VCF file"/>
-        <param name="mutation_regions" type="data" format="bed" label="AMR mutation regions BED file"/>
     </inputs>
     <outputs>
-        <data name="varscan_vcf" format="vcf"/>
+        <data name="output" format="vcf"/>
     </outputs>
     <tests>
         <test>
-            <param name="varscan_raw" value="input1.vcf" ftype="vcf"/>
-            <param name="mutation_regions" value="mutation_regions.bed" ftype="bed"/>
-            <output name="varscan_vcf" value="output1.vcf" ftype="vcf"/>
+            <param name="varscan_raw" value="input.vcf" ftype="vcf"/>
+            <output name="output" value="output.vcf" ftype="vcf"/>
         </test>
     </tests>
     <help>
 **What it does**
 
-Accepts VarScan mpileup2snps file, filters them, and produces a VCF file.
+Accepts VarScan mpileup2snps VCF file, filters it, and produces a filtered VCF file.
     </help>
     <expand macro="citations"/>
 </tool>
-