<tool id="ngsderive_strandedness" name="ngsderive strandedness" version="@TOOL_VERSION@+galaxy@VERSION_SUFFIX@" profile="23.0">
    <description>infers strandedness from RNA-seq BAM files</description>
    <macros>
        <token name="@TOOL_VERSION@">4.0.0</token>
        <token name="@VERSION_SUFFIX@">0</token>
    </macros>
    <requirements>
        <requirement type="package" version="@TOOL_VERSION@">ngsderive</requirement>
    </requirements>
    <command detect_errors="exit_code"><![CDATA[
ln -s '${alignment_input}' input.bam &&
ln -s '${alignment_input.metadata.bam_index}' input.bam.bai &&
ln -s '${gtf_input}' annotation.${gtf_input.ext} &&

ngsderive strandedness
    input.bam
    -g annotation.${gtf_input.ext}
    -n $n_genes
    -m $min_reads_per_gene
    -q $mapq
    $split_by_rg
    > '${output}'
    ]]></command>
    <inputs>
        <param name="alignment_input" type="data" format="bam" label="Input alignment file" help="Aligned paired-end RNA-seq reads in BAM format."/>
        <param name="gtf_input" type="data" format="gtf,gtf.gz" label="Gene annotation file (GTF)" help="Gene model in GTF format. The file will be automatically sorted and indexed if necessary."/>
        <param argument="-n" name="n_genes" type="integer" value="1000" min="1" label="Number of genes to sample" help="Number of random genes to sample for strandedness inference."/>
        <param argument="-m" name="min_reads_per_gene" type="integer" value="10" min="1" label="Minimum reads per gene" help="Minimum number of reads per gene required for inclusion in the analysis."/>
        <param argument="-q" name="mapq" type="integer" value="30" min="0" label="Minimum mapping quality (MAPQ)" help="Minimum MAPQ score for a read to be considered."/>
        <param argument="--split-by-rg" type="boolean" truevalue="--split-by-rg" falsevalue="" checked="false" label="Split results by read group" help="Output one entry per read group in addition to an overall entry."/>
    </inputs>
    <outputs>
        <data name="output" format="tabular" label="${tool.name} on ${on_string}">
            <actions>
                <action name="column_names" type="metadata" default="File,ReadGroup,TotalReads,ForwardPct,ReversePct,Predicted"/>
            </actions>
        </data>
    </outputs>
    <tests>
        <!-- Test forward-stranded data -->
        <test expect_num_outputs="1">
            <param name="alignment_input" value="forward_stranded.bam"/>
            <param name="gtf_input" value="strandedness_test.gtf"/>
            <output name="output">
                <assert_contents>
                    <has_n_columns n="6"/>
                    <has_text text="Stranded-Forward"/>
                </assert_contents>
            </output>
        </test>
        <!-- Test reverse-stranded data -->
        <test expect_num_outputs="1">
            <param name="alignment_input" value="reverse_stranded.bam"/>
            <param name="gtf_input" value="strandedness_test.gtf"/>
            <output name="output">
                <assert_contents>
                    <has_n_columns n="6"/>
                    <has_text text="Stranded-Reverse"/>
                </assert_contents>
            </output>
        </test>
        <!-- Test unstranded data -->
        <test expect_num_outputs="1">
            <param name="alignment_input" value="unstranded.bam"/>
            <param name="gtf_input" value="strandedness_test.gtf"/>
            <output name="output">
                <assert_contents>
                    <has_n_columns n="6"/>
                    <has_text text="Unstranded"/>
                </assert_contents>
            </output>
        </test>
        <!-- Test with gzipped GTF annotation file -->
        <test expect_num_outputs="1">
            <param name="alignment_input" value="reverse_stranded.bam"/>
            <param name="gtf_input" value="strandedness_test.gtf.gz" ftype="gtf.gz"/>
            <output name="output">
                <assert_contents>
                    <has_n_columns n="6"/>
                    <has_text text="Stranded-Reverse"/>
                </assert_contents>
            </output>
        </test>
    </tests>
    <help><![CDATA[
**What it does**

ngsderive strandedness infers the strandedness protocol used to generate RNA-seq data by
analyzing read alignments against a gene model. It can determine whether your data was
generated using a Stranded-Forward, Stranded-Reverse, or Unstranded protocol.

This tool is useful when you have RNA-seq data but are unsure about the library preparation
protocol used. Knowing the correct strandedness is essential for accurate gene expression
quantification.

**How it works**

The tool randomly samples genes from the provided gene model and examines how reads align
to those genes. Based on the proportion of reads mapping in the forward vs reverse orientation,
it classifies the library as:

- **Unstranded**: ~40-60% forward reads
- **Stranded-Forward**: ≥80% forward reads
- **Stranded-Reverse**: ≥80% reverse reads
- **Inconclusive**: Results don't clearly indicate a strandedness type

**Inputs**

- **Alignment file**: Paired-end RNA-seq alignments in BAM format
- **Gene annotation**: GTF file with gene models (gzipped GTF supported)

**Output**

A tabular file with the following columns:

- **File**: Name of the input BAM file
- **ReadGroup**: Read group identifier (or "overall" for combined results)
- **TotalReads**: Number of reads used in the analysis
- **ForwardPct**: Percentage of reads supporting forward strandedness
- **ReversePct**: Percentage of reads supporting reverse strandedness
- **Predicted**: The inferred strandedness (Stranded-Forward, Stranded-Reverse, Unstranded, or Inconclusive)

**Notes**

- Only paired-end reads are currently supported
- For best results, ensure your BAM file has sufficient read depth

For more information, see the `ngsderive documentation <https://stjudecloud.github.io/ngsderive/subcommands/strandedness/>`_.
    ]]></help>
    <citations>
        <citation type="bibtex">
@software{ngsderive,
    author = {{St. Jude Cloud Team}},
    title = {ngsderive: Forensic analysis tool for NGS data},
    url = {https://github.com/stjudecloud/ngsderive},
    year = {2020}
}
        </citation>
    </citations>
</tool>