<tool id="snvmix_somatic_filter" name="Get somatic positions from germline SNVMix output" version="0.12.1-rc1">
  <requirements>
    <requirement type="package">SNVMix</requirement>
  </requirements>
  <description>Takes SNVMix output from a set of specific sites in a bam file from a matched normal sample and the SNV calls from the tumour to identify and report high-confidence somatic mutations</description>
  <command interpreter="python">
    filter_snvmix_somatic.py
      -n $input1
      -t $input2
      -o $output1
      -p $posterior
      -d $nonref_support
  </command>
 <inputs>
   <param name="input1" type="data" format="tabular" label="Select the SNVMix raw output from your normal library" />
   <param name="input2" type="data" format="tabular" label="Select the annotated SNV calls from your tumour library" />
   <param name="posterior" type="float" value="0.999" label="Choose a minimum value for the reference state posterior probability p(AA)" />
   <param name="nonref_support" type="integer" value="2" label="Choose a maximum number of non-reference base calls allowed at a site" />
 </inputs>
  <outputs>
    <data format="tabular" name="output1" label="${tool.name}: Somatic positions" />
  </outputs>
  <help>

**What it does**
Takes the positions identified as SNVs by SNVMix in a tumour sample and compares them to raw SNVMix output from the matched "normal" sequence from the same individual.
The input from the normal must have run SNVMix with the -f (full) option ("SNVMix at selected positions" tool).
The -p option (posterior) is then used to retain only the variants with a strong posterior probability of being "reference" at each position.  
These are the high-confidence somatic SNV calls.  Variants can be further thresholded for enhanced stringency using the -d (nonref_support) option. 
If more than -d non-reference reads exist at the site, the SNV is also considered a "germ line" event and not reported as somatic.  This process returns
only the sites that can be confidently identified as somatic. Thus, with limiting or no coverage at the corresponding position in the germ line,
some true somatic sites may not be identified. 


  </help>
</tool>