view diffdock.xml @ 0:bfba870d3537 draft default tip

planemo upload for repository https://github.com/galaxyproject/tools-iuc/tree/main/tools/diffdock/ commit b0db941a49bdfaa799b10cd68f9b7d8509f608d2

<tool id="diffdock" name="diffdock" version="@TOOL_VERSION@+@VERSION_SUFFIX@" profile="24.2" license="MIT">
    <description>Predict ligand binding poses using DiffDock's diffusion-based docking method</description>
    <macros>
        <import>macros.xml</import>
    </macros>
    <expand macro="requirements" />
    <expand macro="creators" />
    <command detect_errors="aggressive"><![CDATA[
        ln -s /home/appuser/DiffDock/* . &&
        ln -s '$protein_path' protein_path.pdb &&
        ln -s '$ligand_description' ligand_description.sdf &&
        micromamba run -n diffdock python -m inference
        --protein_path protein_path.pdb
        --ligand_description ligand_description.sdf
        --out_dir results
    ]]></command>
    <inputs>
        <param argument="--protein_path" type="data" format="pdb" label="Protein structure" help="3D structure of the protein receptor in PDB format. This is the target for ligand docking."/>
        <param argument="--ligand_description" type="data" format="sdf" label="Ligand molecule" help="Structure of the ligand(s) to be docked in SDF format. Multiple molecules can be included."/>
        <param argument="--samples_per_complex" type="integer" label="Samples per complex" value="10" help="Number of binding pose samples to generate for each protein-ligand complex."/>
        <param argument="--inference_steps" type="integer" label="Total inference steps" value="20" help="Total number of denoising steps to run in the diffusion model."/>
        <param argument="--actual_steps" type="integer" label="Actual denoising steps" value="19" help="Number of denoising steps that are actually performed. Must be less than or equal to inference steps."/>
    </inputs>
    <outputs>
        <collection name="output_collection" type="list" label="${tool.name} on ${on_string}">
            <discover_datasets pattern="__name_and_ext__" directory="results" recurse="true" format="sdf"/>
        </collection>
    </outputs>
    <tests>
    <test expect_num_outputs="1">
        <param name="protein_path" value="1a0q_protein_processed.pdb"/>
        <param name="ligand_description" value="1a0q_ligand.sdf"/>
        <param name="samples_per_complex" value="10"/>
        <param name="inference_steps" value="20"/>
        <param name="actual_steps" value="19"/>
        <output_collection name="output_collection" type="list">
            <element name="rank1">
                <assert_contents>
                    <has_n_lines n="52"/>
                    <has_line line="1a0q_ligand"/>
                    <has_line line=" 23 23  0  0  0  0  0  0  0  0999 V2000"/>
                </assert_contents>
            </element>
        </output_collection>
    </test>
    </tests>
    <help><![CDATA[
DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking
------------------------------------------------------------------

DiffDock is a novel deep learning-based docking method using diffusion models to predict ligand binding poses.

For more information, visit the `DiffDock GitHub repository <https://github.com/gcorso/DiffDock>`_

**License**

* `MIT <https://raw.githubusercontent.com/gcorso/DiffDock/refs/heads/main/LICENSE>`_

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