<tool id="snpSift_extractFields" name="SnpSift Extract Fields" version="@WRAPPER_VERSION@.galaxy0">
    <options sanitize="False" />
    <description>from a VCF file into a tabular file</description>
    <macros>
        <import>snpSift_macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <expand macro="stdio" />
    <expand macro="version_command" />
    <command><![CDATA[
@CONDA_SNPSIFT_JAR_PATH@ &&
cat '${input}'
#if $one_effect_per_line:
    | perl "\$SNPSIFT_JAR_PATH/scripts/vcfEffOnePerLine.pl"
#end if
| SnpSift -Xmx6G extractFields
#if $separator:
    -s '${separator}'
#end if
#if $empty_text:
    -e '${empty_text}'
#end if
-
#echo ' '.join(['"%s"' % x for x in $extract.split()])
> '${output}'
    ]]></command>
    <inputs>
        <param name="input" type="data" format="vcf" label="Variant input file in VCF format"/>
        <param name="extract" type="text" label="Fields to extract" value="CHROM POS ID REF ALT FILTER" help="Separated by spaces. See help below for an explanation" />
        <param name="one_effect_per_line" type="boolean" truevalue="yes" falsevalue="no" checked="false" label="One effect per line" help="When variants have more than one effect, lists one effect per line, while all other parameters in the line are repeated across mutiple lines" />
        <param name="separator" type="text" value="" label="multiple field separator" help="Separate multiple fields in one column with this character, e.g. a comma, rather than a column for each of the multiple values" argument="-s" />
        <param name="empty_text" type="text" value="" label="empty field text" help="Represent empty fields with this value, rather than leaving them blank" argument="-e"/>
    </inputs>
    <outputs>
        <data name="output" format="tabular" />
    </outputs>
    <tests>
        <test>
            <param name="input" ftype="vcf" value="test_rmInfo.vcf"/>
            <param name="extract" value="CHROM POS REF ALT EFF[*].EFFECT"/>
            <output name="output">
                <assert_contents>
                    <has_text text="INTRAGENIC" />
                    <not_has_text text="DOWNSTREAM,INTRAGENIC,INTRON,UTR_3_PRIME" />
                </assert_contents>
            </output>
        </test>
        <test>
            <param name="input" ftype="vcf" value="test_rmInfo.vcf"/>
            <param name="extract" value="CHROM POS REF ALT EFF[*].EFFECT"/>
            <param name="separator" value=","/>
            <output name="output">
                <assert_contents>
                    <has_text text="DOWNSTREAM,INTRAGENIC,INTRON,UTR_3_PRIME" />
                </assert_contents>
            </output>
        </test>
        <test>
            <param name="input" ftype="vcf" value="extFields_test3_in.vcf"/>
            <param name="extract" value="CHROM POS ID REF ALT FILTER ANN[*].EFFECT"/>
            <param name="one_effect_per_line" value="true"/>
            <output name="output" value="extFields_test3_out.vcf"/>
        </test>
    </tests>
    <help><![CDATA[
**What is does**

`SnpSift Extract Fields <http://snpeff.sourceforge.net/SnpSift.html#Extract>`_ selects columns from a VCF dataset into a Tab-delimited format.

------

.. class:: infomark

**How to know which fields to extract?**

A VCF dataset contains mandatory fields as well as optional fields. Mandatory fields are required by `VCF specifications <https://samtools.github.io/hts-specs/VCFv4.2.pdf>`_ and present in any valid VCF dataset. The **Fields to extract** input box of the tool above is already pre-filled with names of mandatory fields.

To know what other fields are available in a given VCF file simply look at its header. `INFO` and `FORMAT` lines will contain description of existing fields. For example, if you see a line:

##INFO=<ID=NS,Number=1,Type=Integer,Description="Number of samples with data">

you can use *NS* as the field name.

------

**Dealing with field generated with SnpEff**

The current version of `SnpEff <http://snpeff.sourceforge.net/SnpEff_manual.html>`_ produces so called *ANN* fields::

    "ANN[*].ALLELE" (alias GENOTYPE)
    "ANN[*].EFFECT" (alias ANNOTATION): Effect in Sequence ontology terms (e.g. 'missense_variant', 'synonymous_variant', 'stop_gained', etc.)
    "ANN[*].IMPACT" { HIGH, MODERATE, LOW, MODIFIER }
    "ANN[*].GENE" Gene name (e.g. 'PSD3')
    "ANN[*].GENEID" Gene ID
    "ANN[*].FEATURE"
    "ANN[*].FEATUREID" (alias TRID: Transcript ID)
    "ANN[*].BIOTYPE" Biotype, as described by the annotations (e.g. 'protein_coding')
    "ANN[*].RANK" Exon or Intron rank (i.e. exon number in a transcript)
    "ANN[*].HGVS_C" (alias HGVS_DNA, CODON): Variant in HGVS (DNA) notation
    "ANN[*].HGVS_P" (alias HGVS, HGVS_PROT, AA): Variant in HGVS (protein) notation
    "ANN[*].CDNA_POS" (alias POS_CDNA)
    "ANN[*].CDNA_LEN" (alias LEN_CDNA)
    "ANN[*].CDS_POS" (alias POS_CDS)
    "ANN[*].CDS_LEN" (alias LEN_CDS)
    "ANN[*].AA_POS" (alias POS_AA)
    "ANN[*].AA_LEN" (alias LEN_AA)
    "ANN[*].DISTANCE"
    "ANN[*].ERRORS" (alias WARNING, INFOS)

Older versions produced *EFF* fields::

    "EFF[*].EFFECT"
    "EFF[*].IMPACT"
    "EFF[*].FUNCLASS"
    "EFF[*].CODON"
    "EFF[*].AA"
    "EFF[*].AA_LEN"
    "EFF[*].GENE"
    "EFF[*].BIOTYPE"
    "EFF[*].CODING"
    "EFF[*].TRID"
    "EFF[*].RANK"

In addition there are *LOF* and *NMD* fields::

    "LOF[*].GENE"
    "LOF[*].GENEID"
    "LOF[*].NUMTR"
    "LOF[*].PERC"

    "NMD[*].GENE"
    "NMD[*].GENEID"
    "NMD[*].NUMTR"
    "NMD[*].PERC"

To find our whether your VCF contains *ANN* or *EFF* annotations simply look at its header.

-----

**Usage examples**

*Extracting chromosome, position, ID and allele frequency from a VCF file*:

**CHROM POS ID AF**

The result will look something like::

    #CHROM        POS        ID            AF
    1             69134                    0.086
    1             69496      rs150690004   0.001

*Extracting genotype fields*:

**CHROM POS ID THETA GEN[0].GL[1] GEN[1].GL GEN[3].GL[*] GEN[*].GT**

This means to extract:

- CHROM POS ID: regular fields (as in the previous example)
- THETA : This one is from INFO
- GEN[0].GL[1] : Second likelihood from first genotype
- GEN[1].GL : The whole GL field (all entries without separating them)
- GEN[3].GL[*] : All likelihoods form genotype 3 (this time they will be tab separated, as opposed to the previous one).
- GEN[*].GT : Genotype subfields (GT) from ALL samples (tab separated).

The result will look something like::

    #CHROM  POS     ID              THETA   GEN[0].GL[1]    GEN[1].GL               GEN[3].GL[*]            GEN[*].GT
    1       10583   rs58108140      0.0046  -0.47           -0.24,-0.44,-1.16       -0.48   -0.48   -0.48   0|0     0|0     0|0     0|1     0|0     0|1     0|0     0|0     0|1
    1       10611   rs189107123     0.0077  -0.48           -0.24,-0.44,-1.16       -0.48   -0.48   -0.48   0|0     0|1     0|0     0|0     0|0     0|0     0|0     0|0     0|0
    1       13302   rs180734498     0.0048  -0.58           -2.45,-0.00,-5.00       -0.48   -0.48   -0.48   0|0     0|1     0|0     0|0     0|0     1|0     0|0     0|1     0|0

*Extracting fields with multiple values*:
    (notice that there are multiple effect columns per line because there are multiple effects per variant)

**CHROM POS REF ALT ANN[*].EFFECT**

The result will look something like::

    #CHROM  POS REF ALT ANN[*].EFFECT
    22  17071756    T   C   3_prime_UTR_variant downstream_gene_variant
    22  17072035    C   T   missense_variant    downstream_gene_variant
    22  17072258    C   A   missense_variant    downstream_gene_variant

*Extracting fields with multiple values using a comma as a multiple field separator:*

**CHROM POS REF ALT ANN[*].EFFECT ANN[*].HGVS_P**

The result will look something like::

    #CHROM  POS REF ALT ANN[*].EFFECT   ANN[*].HGVS_P
    22  17071756    T   C   3_prime_UTR_variant,downstream_gene_variant .,.
    22  17072035    C   T   missense_variant,downstream_gene_variant    p.Gly469Glu,.
    22  17072258    C   A   missense_variant,downstream_gene_variant    p.Gly395Cys,.

*Extracting fields with multiple values, one effect per line:*

**CHROM POS REF ALT ANN[*].EFFECT**

The result will look something like::

    #CHROM  POS REF ALT ANN[*].EFFECT
    22  17071756    T   C   3_prime_UTR_variant
    22  17071756    T   C   downstream_gene_variant
    22  17072035    C   T   missense_variant
    22  17072035    C   T   downstream_gene_variant
    22  17072258    C   A   missense_variant
    22  17072258    C   A   downstream_gene_variant

@EXTERNAL_DOCUMENTATION@
- http://snpeff.sourceforge.net/SnpSift.html#Extract
    ]]></help>
    <expand macro="citations" />
</tool>