Merge pull request #116 from nf-core/implement-neo-subworkflow

README.md

@@ -57,6 +57,7 @@ The following processes and tools can be run with `oncoanalyser`:
 - HRD status prediction (`CHORD`)
 - Mutational signature fitting (`Sigs`)
 - Tissue of origin prediction (`CUPPA`)
+- Neoepitope prediction (`Neo`)
 - Report generation (`ORANGE`, `linxreport`)
 
 ## Usage

conf/hmf_data.config

@@ -20,8 +20,12 @@ params {
             isofox_gc_ratios              = 'rna_pipeline/read_100_exp_gc_ratios.csv'
             // LILAC
             lilac_resources               = 'dna_pipeline/immune/'
+            // Neo
+            neo_resources                 = 'neo/binding/'
             // ORANGE
             cohort_mapping                = 'orange/cohort_mapping.tsv'
+            // Hartwig cohort RNA data
+            cohort_tpm_medians            = 'neo/tpm_cohort/hmf_tpm_medians.37.csv'
             cohort_percentiles            = 'orange/cohort_percentiles.tsv'
             alt_sj_distribution           = 'rna_pipeline/isofox.hmf_3444.alt_sj_cohort.37.csv'
             gene_exp_distribution         = 'rna_pipeline/isofox.hmf_3444.gene_distribution.37.csv'
@@ -72,8 +76,12 @@ params {
             isofox_gc_ratios              = 'rna_pipeline/read_100_exp_gc_ratios.csv'
             // LILAC
             lilac_resources               = 'dna_pipeline/immune/'
+            // Neo
+            neo_resources                 = 'neo/binding/'
             // ORANGE
             cohort_mapping                = 'orange/cohort_mapping.tsv'
+            // Hartwig cohort RNA data
+            cohort_tpm_medians            = 'neo/tpm_cohort/hmf_tpm_medians.38.csv'
             cohort_percentiles            = 'orange/cohort_percentiles.tsv'
             alt_sj_distribution           = 'rna_pipeline/isofox.hmf_3444.alt_sj_cohort.38.csv'
             gene_exp_distribution         = 'rna_pipeline/isofox.hmf_3444.gene_distribution.38.csv'

conf/modules.config

@@ -218,6 +218,34 @@ process {
         ]
     }
 
+    withName: 'NEO_(?:SCORER|FINDER)' {
+        ext.jarPath = '/opt/neo/neo.jar'
+    }
+
+    withName: 'NEO_SCORER' {
+        publishDir = [
+            path: { "${params.outdir}" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : "${meta.key}/neo/scorer/" },
+        ]
+    }
+
+    withName: '.*:NEO_PREDICTION:ANNOTATE_FUSIONS' {
+        publishDir = [
+            path: { "${params.outdir}" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : "${meta.key}/neo/annotated_fusions/${filename}" },
+        ]
+    }
+
+    withName: 'NEO_FINDER' {
+        publishDir = [
+            path: { "${params.outdir}" },
+            mode: params.publish_dir_mode,
+            saveAs: { filename -> filename.equals('versions.yml') ? null : "${meta.key}/neo/finder/" },
+        ]
+    }
+
     withName: 'CUPPA' {
         publishDir = [
             path: { "${params.outdir}" },

docs/output.md

@@ -71,6 +71,8 @@ output/
   - [Sigs](#sigs) - Mutational signature fitting
 - [Tissue of origin prediction](#tissue-of-origin-prediction)
   - [CUPPA](#cuppa) - Tissue of origin prediction
+- [Neoepitope prediction](#neoepitope-prediction)
+  - [Neo](#neo) - Neoepitope prediction
 - [Report generation](#report-generation)
   - [ORANGE](#orange) - Key results summary
   - [linxreport](#linxreport) - Interactive LINX report
@@ -473,6 +475,23 @@ signatures to tumor sample data.
 [CUPPA](https://github.com/hartwigmedical/hmftools/tree/master/cuppa) predicts tissue of origin for a given tumor sample
 using DNA and/or RNA features generated by upstream hmftools components.
 
+### Neoepitope prediction
+
+#### Neo
+
+<details markdown="1">
+<summary>Output files</summary>
+
+- `<group_id>/neo/`
+  - `<tumor_dna_id>.neo.neo_data.tsv`: Neoepitope candidates.
+  - `<tumor_dna_id>.neo.neoepitope.tsv`: LINX fusion neoepitopes.
+  - `<tumor_dna_id>.neo.peptide_scores.tsv`: Peptide binding likelihood and scoring.
+
+</details>
+
+[Neo](https://github.com/hartwigmedical/hmftools/tree/master/neo) builds comprehensive neoepitope predictions from DNA
+data with additional annotations made using RNA data.
+
 ### Report generation
 
 #### ORANGE

lib/Constants.groovy

@@ -47,6 +47,7 @@ class Constants {
         LILAC,
         LINX,
         MARKDUPS,
+        NEO,
         ORANGE,
         PAVE,
         PURPLE,

lib/Processes.groovy

@@ -7,7 +7,17 @@ import Utils
 class Processes {
 
     public static getRunStages(include, exclude, manual_select, log) {
-        def processes = manual_select ? [] : Constants.Process.values().toList()
+
+        // Get default processes
+        // NOTE(SW): currently set all except Neo to run by default; Process.NEO excluded to be more concise in code
+        def processes
+        if (manual_select) {
+            processes = []
+        } else {
+            processes = Constants.Process.values().toList()
+            processes.remove(Constants.Process.NEO)
+        }
+
         def include_list = this.getProcessList(include, log)
         def exclude_list = this.getProcessList(exclude, log)
         this.checkIncludeExcludeList(include_list, exclude_list, log)

modules/local/linx/somatic/main.nf

@@ -38,6 +38,7 @@ process LINX_SOMATIC {
         -known_fusion_file ${known_fusion_data} \\
         -driver_gene_panel ${driver_gene_panel} \\
         -write_vis_data \\
+        -write_neo_epitopes \\
         -output_dir linx_somatic/
 
     cat <<-END_VERSIONS > versions.yml

modules/local/neo/Dockerfile

@@ -0,0 +1,22 @@
+FROM mambaorg/micromamba:0.24.0
+
+USER root
+
+RUN \
+    apt-get update && \
+    apt-get install -y procps wget && \
+    apt-get clean && \
+    rm -rf /var/lib/apt/lists/*
+
+RUN \
+    mkdir -p /opt/neo/ && \
+    wget -O /opt/neo/neo.jar 'https://github.com/hartwigmedical/hmftools/releases/download/neo-v1.2_beta/neo_v1.2_beta.jar'
+
+USER mambauser
+
+RUN \
+    micromamba install -y -n base -c bioconda -c conda-forge \
+        'openjdk >=8' && \
+    micromamba clean --all --yes
+
+ENV PATH="/opt/conda/bin:/opt/conda/condabin:${PATH}"

modules/local/neo/annotate_fusions/environment.yml

@@ -0,0 +1,7 @@
+name: isofox
+channels:
+  - conda-forge
+  - bioconda
+  - defaults
+dependencies:
+  - bioconda::hmftools-isofox=1.7.1

modules/local/neo/annotate_fusions/main.nf

@@ -0,0 +1,56 @@
+process ANNOTATE_FUSIONS {
+    tag "${meta.id}"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "${ workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container ?
+        'https://depot.galaxyproject.org/singularity/hmftools-isofox:1.7.1--hdfd78af_0' :
+        'biocontainers/hmftools-isofox:1.7.1--hdfd78af_0' }"
+
+    input:
+    tuple val(meta), path(neo_finder_dir), path(bam), path(bai)
+    val read_length
+    path genome_fasta
+    val genome_ver
+    path genome_fai
+    path ensembl_data_resources
+
+    output:
+    tuple val(meta), path('*isf.neoepitope.tsv'), emit: annotated_fusions
+    path 'versions.yml'                         , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+
+    """
+    mkdir -p isofox/
+
+    isofox \\
+        -Xmx${Math.round(task.memory.bytes * 0.95)} \\
+        ${args} \\
+        -sample ${meta.sample_id} \\
+        -bam_file ${bam} \\
+        -functions NEO_EPITOPES \\
+        -neo_dir ${neo_finder_dir} \\
+        -read_length ${read_length} \\
+        -ref_genome ${genome_fasta} \\
+        -ref_genome_version ${genome_ver} \\
+        -ensembl_data_dir ${ensembl_data_resources} \\
+        -threads ${task.cpus} \\
+        -output_dir ./
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        isofox: \$(isofox -version | sed 's/^.* //')
+    END_VERSIONS
+    """
+
+    stub:
+    """
+    touch ${meta.sample_id}.isf.neoepitope.tsv
+    echo -e '${task.process}:\\n  stub: noversions\\n' > versions.yml
+    """
+}

modules/local/neo/annotate_fusions/meta.yml

@@ -0,0 +1,63 @@
+name: annotate_fusions
+description: Annotate neoeptitopes with RNA fusion data
+keywords:
+  - neoepitopes
+  - rna
+  - rnaseq
+tools:
+  - isofox:
+      description: Characterises and counts gene, transcript features
+      homepage: https://github.com/hartwigmedical/hmftools/tree/master/isofox
+      documentation: https://github.com/hartwigmedical/hmftools/tree/master/isofox
+      licence: ["GPL v3"]
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [id: 'sample_id']
+  - neo_finder_dir:
+      type: directory
+      description: Neo Finder directory
+  - bam:
+      type: file
+      description: BAM file
+      pattern: "*.{bam}"
+  - bai:
+      type: file
+      description: BAI file
+      pattern: "*.{bai}"
+  - read_length:
+      type: integer
+      description: Read length
+  - genome_fasta:
+      type: file
+      description: Reference genome assembly FASTA file
+      pattern: "*.{fa,fasta}"
+  - genome_ver:
+      type: string
+      description: Reference genome version
+  - genome_fai:
+      type: file
+      description: Reference genome assembly fai file
+      pattern: "*.{fai}"
+  - ensembl_data_resources:
+      type: directory
+      description: HMF ensembl data resources directory
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [id: 'sample_id']
+  - annotated_fusions:
+      type: file
+      description: Annotated neoepitopes file
+      pattern: "*.{tsv}"
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+authors:
+  - "@scwatts"
+  - "@charlesshale"

modules/local/neo/finder/main.nf

@@ -0,0 +1,52 @@
+process NEO_FINDER {
+    tag "${meta.id}"
+    label 'process_low'
+
+    container 'docker.io/scwatts/neo:1.2_beta--1'
+
+    input:
+    tuple val(meta), path(purple_dir), path(linx_annotation_dir)
+    path genome_fasta
+    val genome_ver
+    path genome_fai
+    path ensembl_data_resources
+
+    output:
+    tuple val(meta), path('neo_finder/'), emit: neo_finder_dir
+    path 'versions.yml'                 , emit: versions
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def args = task.ext.args ?: ''
+
+    """
+    mkdir -p neo_finder/
+
+    java \\
+        -Xmx${Math.round(task.memory.bytes * 0.95)} \\
+        -jar ${task.ext.jarPath} \\
+            ${args} \\
+            -sample ${meta.sample_id} \\
+            -linx_dir ${linx_annotation_dir} \\
+            -somatic_vcf ${purple_dir}/${meta.sample_id}.purple.somatic.vcf.gz \\
+            -ref_genome ${genome_fasta} \\
+            -ref_genome_version ${genome_ver} \\
+            -ensembl_data_dir ${ensembl_data_resources} \\
+            -log_debug \\
+            -output_dir neo_finder/
+
+    cat <<-END_VERSIONS > versions.yml
+    "${task.process}":
+        neo: \$(java -jar ${task.ext.jarPath} -version | sed 's/^.* //')
+    END_VERSIONS
+    """
+
+    stub:
+    """
+    mkdir -p neo_finder/
+    echo -e '${task.process}:\\n  stub: noversions\\n' > versions.yml
+    """
+}
+

modules/local/neo/finder/meta.yml

@@ -0,0 +1,52 @@
+name: neo_finder
+description: Identify candidate neoeptitops
+keywords:
+  - neoepitopes
+tools:
+  - neo:
+      description: Predict and score neoepitopes
+      homepage: https://github.com/hartwigmedical/hmftools/tree/master/neo
+      documentation: https://github.com/hartwigmedical/hmftools/tree/master/neo
+      licence: ["GPL v3"]
+input:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [id: 'sample_id']
+  - purple_dir:
+      type: directory
+      description: PURPLE output directory
+  - linx_annotation_dir:
+      type: directory
+      description: LINX somatic annotation output directory
+  - genome_fasta:
+      type: file
+      description: Reference genome assembly FASTA file
+      pattern: "*.{fa,fasta}"
+  - genome_ver:
+      type: string
+      description: Reference genome version
+  - genome_fai:
+      type: file
+      description: Reference genome assembly fai file
+      pattern: "*.{fai}"
+  - ensembl_data_resources:
+      type: directory
+      description: HMF ensembl data resources directory
+output:
+  - meta:
+      type: map
+      description: |
+        Groovy Map containing sample information
+        e.g. [id: 'sample_id']
+  - neo_finder_dir:
+      type: directory
+      description: Neo Finder output directory
+  - versions:
+      type: file
+      description: File containing software versions
+      pattern: "versions.yml"
+authors:
+  - "@scwatts"
+  - "@charlesshale"