<?xml version="1.0" encoding="UTF-8"?>
<tool name="CollectHsMetrics" id="picard_CollectHsMetrics" version="@TOOL_VERSION@">
    <description>compute metrics about datasets generated through hybrid-selection (e.g. exome)</description>
    <macros>
        <import>picard_macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <command detect_errors="exit_code"><![CDATA[
@java_options@
@symlink_element_identifier@

picard
CollectHsMetrics
INPUT='$escaped_element_identifier'
OUTPUT='${outFile}'
MINIMUM_MAPPING_QUALITY=${minimum_mapping_quality}
MINIMUM_BASE_QUALITY=${minimum_base_quality}
COVERAGE_CAP=${coverage_cap}
CLIP_OVERLAPPING_READS=${clip_overlapping_reads}
BAIT_INTERVALS='${baitFile}'
TARGET_INTERVALS='${targetFile}'
VALIDATION_STRINGENCY=${validation_stringency}
QUIET=true
VERBOSITY=ERROR
    ]]></command>
    <inputs>
        <param name="inputFile" type="data" format="sam,bam" label="Select SAM/BAM dataset or dataset collection" help="If empty, upload or import a SAM/BAM dataset" />
        <param name="baitFile" type="data" format="picard_interval_list" label="An interval list file that contains the locations of the baits used" help="If empty, upload or convert from BED" />
        <param name="targetFile" type="data" format="picard_interval_list" label="An interval list file that contains the locations of the targets" help="If empty, upload or convert from BED" />
        <param name="minimum_mapping_quality" type="integer" value="20" label="Minimum mapping quality for a read to contribute coverage" help="MINIMUM_MAPPING_QUALITY; default=20" />
        <param name="minimum_base_quality" type="integer" value="20" label="Minimum base quality for a base to contribute coverage" help="MINIMUM_BASE_QUALITY; default=20" />
        <param name="coverage_cap" type="integer" value="200" label="Treat bases with coverage exceeding this value as if they had coverage at this value" help="COVERAGE_CAP; default=200" />
        <param name="clip_overlapping_reads" type="boolean" label="Clip overlapping reads" checked="true" truevalue="true" falsevalue="false" help="CLIP_OVERLAPPING_READS; default=true" />
        <expand macro="VS" />
    </inputs>
    <outputs>
        <data name="outFile" format="txt" label="${tool.name} on ${on_string}: Summary data" />
    </outputs>
    <tests>
        <test>
            <param name="inputFile" value="picard_CollectHsMetrics.bam" ftype="bam" />
            <param name="baitFile" value="picard_CollectHsMetrics.baits" ftype="picard_interval_list" />
            <param name="targetFile" value="picard_CollectHsMetrics.target" ftype="picard_interval_list" />
            <param name="minimum_mapping_quality" value="20" />
            <param name="minimum_base_quality" value="20" />
            <param name="coverage_cap" value="200" />
            <param name="clip_overlapping_reads" value="true" />
            <output name="outFile" file="picard_CollectHsMetrics_test1.tab" lines_diff="6" />
        </test>
    </tests>
    <help><![CDATA[
.. class:: infomark

**Purpose**

Computes a number of metrics that are useful for evaluating coverage and performance of datasets generated through hybrid-selection. 
Hybrid-selection (HS) is the most commonly used technique to capture exon-specific sequences for targeted sequencing experiments 
such as exome sequencing.

@dataset_collections@

-----

.. class:: warningmark

**Obtaining bait and target interval files in Picard interval_list format**

This tool requires an aligned SAM or BAM file as well as bait and target interval files in Picard interval_list format. 
You should use the bait and interval files that correspond to the capture kit that was used to generate the capture libraries
for sequencing, which can generally be obtained from the kit manufacturer.
If the baits and target intervals are provided in BED format, you can convert them to the Picard interval_list format 
using Picard's **BedToIntervalList** tool, which will also add the required SAM style header.

@description@

  MINIMUM_MAPPING_QUALITY=Integer
  MQ=Integer                    Minimum mapping quality for a read to contribute coverage.  Default value: 20.

  MINIMUM_BASE_QUALITY=Integer
  Q=Integer                     Minimum base quality for a base to contribute coverage.  Default value: 20.

  COVERAGE_CAP=Integer
  CAP=Integer                   Treat bases with coverage exceeding this value as if they had coverage at this value.
                                Default value: 200.
  CLIP_OVERLAPPING_READS=Boolean
                                If true, clip overlapping reads, false otherwise. Default value: true.

@more_info@
    ]]></help>
    <expand macro="citations" />
</tool>