view snpSift_filter.xml @ 8:5fab4f81391d draft default tip

"planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/master/tool_collections/snpsift/snpsift commit df8a21df77acffe40c0bc0fe0409ca1b529cd7fc"

<tool id="snpSift_filter" name="SnpSift Filter" version="@WRAPPER_VERSION@.galaxy1">
    <description>Filter variants using arbitrary expressions</description>
    <macros>
        <import>snpSift_macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <expand macro="stdio" />
    <expand macro="version_command" />
    <command><![CDATA[
SnpSift -Xmx6G filter -f '$input' -e '$exprFile' $inverse
#if str($filter_expression.type) == 'complex':
    #for $set_file in $filter_expression.set:
        --set $set_file
    #end for
#end if
#if $filtering.mode == 'set_filter':
    --filterID $filtering.filter_id
#elif $filtering.mode == 'remove_filter':
    --rmFilter $filtering.rm_filter
#elif $filtering.mode == 'add_filter':
    --addFilter $filtering.add_filter
#end if
> '$output'
    ]]></command>
    <configfiles>
        <configfile name="exprFile">
$filter_expression.expr#slurp
        </configfile>
    </configfiles>
    <inputs>
        <param name="input" type="data" format="vcf" label="Input variant list in VCF format"/>
        <conditional name="filter_expression">
            <param name="type" type="select"
            label="Type of filter expression">
                <option value="simple">Simple expression</option>
                <option value="complex">Expression using value set(s)</option>
            </param>
            <when value="simple">
                <param name="expr" type="text" label="Filter criteria"
                help="Need help? See the tool help below for some examples.">
                    <sanitizer sanitize="False"/>
                </param>
            </when>
            <when value="complex">
                <param name="expr" type="text" label="Filter criteria"
                help="Need help? See the tool help below for some examples.">
                    <sanitizer sanitize="False"/>
                </param>
                <param name="set" type="data" format="txt" multiple="true" optional="False"
                label="Set value source"
                help="Select one or more datasets for construction of value sets. The datasets are supposed to specify one value per line. See also: the help section on Sets below." />
            </when>
        </conditional>
        <param argument="--inverse" type="boolean" truevalue="--inverse" falsevalue="" checked="false" label="Invert filter" help="Select variants that do not match the filter expression." />
        <conditional name="filtering">
            <param name="mode" type="select" label="Filter mode">
                <option value="entries" selected="true">Retain selected variants, remove others</option>
                <option value="set_filter">Add a value to FILTER field of non-selected variants</option>
                <option value="remove_filter">Remove a value from FILTER field of selected variants</option>
                <option value="add_filter">Add a value to the FILTER field of selected variants</option>
            </param>
            <when value="entries" />
            <when value="set_filter">
                <param argument="--filterID" name="filter_id" type="text" optional="false"
                label="Value to add to FILTER field of non-selected variants"
                help="The value provided here will be added to the FILTER field of variants that are NOT selected by the expression above. Selected variants, in contrast, will have their FILTER field cleared and set to PASS." />
            </when>
            <when value="remove_filter">
                <param argument="--rmFilter" name="rm_filter" type="text" optional="false"
                label="Value to remove from FILTER field of selected variants"
                help="The FILTER field of variants that are selected by the expression above will be checked for existence of the specified value. When the value is found it will be removed, but other values will be preserved. If the removed value was the only value in a FILTER field, that field will be set to the '.' (missing) value." />
            </when>
            <when value="add_filter">
                <param argument="--addFilter" name="add_filter" type="text" optional="false"
                label="Value to add to FILTER field of selected variants"
                help="The value provided here will be added to the FILTER field of variants that are selected by the expression above. USE WITH CARE: The tool will NOT add a corresponding FILTER entry to the VCF header of the output, i.e. will produce output that violates the VCF format specification. To avoid issues with downstream tools it may be necessary to add such a header line using standard text processing tools before proceeding." />
            </when>
        </conditional>
    </inputs>
    <outputs>
        <data name="output" format="vcf" />
    </outputs>
    <tests>
        <test>
        <param name="input" ftype="vcf" value="test01.vcf"/>
        <param name="expr" value="QUAL >= 50"/>
        <param name="mode" value="entries"/>
        <output name="output">
            <assert_contents>
            <has_text text="28837706" />
            <not_has_text text="NT_166464" />
            </assert_contents>
        </output>
        </test>
        <test>
        <param name="input" ftype="vcf" value="test01.vcf"/>
        <param name="expr" value="(CHROM = '19')"/>
        <param name="mode" value="entries"/>
        <output name="output">
            <assert_contents>
            <has_text text="3205820" />
            <not_has_text text="NT_16" />
            </assert_contents>
        </output>
        </test>
        <test>
        <param name="input" ftype="vcf" value="test01.vcf"/>
        <param name="expr" value="(POS &gt;= 20175) &amp; (POS &lt;= 35549)"/>
        <param name="mode" value="entries"/>
        <output name="output">
            <assert_contents>
            <has_text text="20175" />
            <has_text text="35549" />
            <has_text text="22256" />
            <not_has_text text="18933" />
            <not_has_text text="37567" />
            </assert_contents>
        </output>
        </test>
        <test>
        <param name="input" ftype="vcf" value="test01.vcf"/>
        <param name="expr" value="( DP &gt;= 5 )"/>
        <param name="mode" value="entries"/>
        <output name="output">
            <assert_contents>
            <has_text text="DP=5;" />
            <has_text text="DP=6;" />
            <not_has_text text="DP=1;" />
            </assert_contents>
        </output>
        </test>
        <test>
        <param name="input" ftype="vcf" value="test01.vcf"/>
        <conditional name="filter_expression">
            <param name="type" value="complex" />
            <param name="expr" value="( POS in SET[0] ) | ( POS in SET[1] )" />
            <param name="set" ftype="txt" value="test_set1.txt,test_set2.txt" />
        </conditional>
        <param name="mode" value="entries"/>
        <output name="output">
            <assert_contents>
                <has_text text="NT_166464&#009;7268&#009;" />
                <has_text text="NT_166464&#009;7283&#009;" />
                <has_text text="NT_166464&#009;7335&#009;" />
                <has_text text="NT_166480&#009;12474&#009;" />
                <has_text text="NT_166480&#009;12483&#009;" />
                <not_has_text text="NT_166464&#009;7258&#009;" />
                <not_has_text text="NT_166452&#009;16693&#009;" />
            </assert_contents>
        </output>
        </test>
    </tests>
    <help><![CDATA[
**SnpSift filter**

This tool provides a flexible solution for filtering the variants in a VCF input dataset through the use of arbitrary, possibly rather complex expressions.

Some examples:

- *I want just the variants from the second million bases of chr1*::

      ( CHROM = 'chr1' ) & ( POS > 1000000 ) & ( POS < 2000000 )

- *Filter value is either 'PASS' or it is missing*::

      (FILTER = 'PASS') | ( na FILTER )

- *Variants that have either a QUAL score above 30, or are indel variants, or for which at least two samples have a heterozygous genotype called*:: 

      (QUAL > 30) | (exists INDEL) | ( countHet() > 2 )

- *Variants that are supported by at least 10 reads (as calculated from the DP4 attribute in the INFO field through zero-based index-access to the multiple values)*::

      (DP4[2] + DP4[3] >= 10)

----

Sets:

The tool can construct sets of values for use in expressions from text files listing one value per line. Variants can then be filtered based on whether a given field in the variant record has a value that's contained in a set. For example, the expression::

      ( ID in SET[2] )

would filter variants based on whether their ID field value appears in the set parsed from the third dataset used for set construction (the first set can be addressed with index ``[0]``, the second with index ``[1]``, and so on).

----

Genotype-based filtering:

Genotypes of specific samples can be accessed via zero-based indexing or via sample names.

- *I want to keep samples where the genotype for the first sample is homozygous variant and the genotype for the second sample is reference*::

      (isHom( GEN[0] ) & isVariant( GEN[0] ) & isRef( GEN[1] ))

----

Filtering based on SnpEff annotations (``ANN`` or ``EFF`` fields):

- *I want to filter lines with an ANN annotation EFFECT of 'frameshift_variant' ( for vcf files using Sequence Ontology terms )*::

      ( ANN[*].EFFECT has 'frameshift_variant' )

  .. class:: infomark

  According to the specification, there can be more than one EFFECT separated by ``&`` (e.g. ``'missense_variant&splice_region_variant'``), thus using the ``has`` operator is better than using the equality operator (``=``). For instance, ``'missense_variant&splice_region_variant' = 'missense_variant'`` is false, whereas ``'missense_variant&splice_region_variant' has 'missense_variant'`` is true.

- *I want to filter lines with an EFF of 'FRAME_SHIFT' ( for vcf files using Classic Effect names )*::

      ( EFF[*].EFFECT = 'FRAME_SHIFT' )


.. class:: infomark

For information regarding HGVS and Sequence Ontology terms versus classic names:

- https://pcingola.github.io/SnpEff/se_commandline/ for the options: ``-classic``, ``-hgvs``, and ``-sequenceOntology``
- https://pcingola.github.io/SnpEff/se_inputoutput/#effect-prediction-details for the table containing the classic name and sequence onology term for each effect

@EXTERNAL_DOCUMENTATION@
- https://pcingola.github.io/SnpEff/ss_filter/

The second link in particular has further details and more examples about the tool's expression syntax.
    ]]></help>
    <expand macro="citations" />
</tool>