15 implement cancer predisposition sequencing reporter

CHANGELOG.md

@@ -17,6 +17,7 @@ Team from October 2025 nf-core Hackathon: @georgiakes, @petanska, @maxibor, @eol
 - [#58](https://github.com/nf-core/variantprioritization/pull/58) - Defined a function that extract the variantcaller from the vcf (@yussab)
 - [#59](https://github.com/nf-core/variantprioritization/pull/59) - Enable conda usage for the entire pipeline (@yussab)
 - [#63](https://github.com/nf-core/variantprioritization/pull/63) - Restructure local modules (@famosab)
+- [#68](https://github.com/nf-core/variantprioritization/pull/68) - Cancer Predisposition Reporter (@yussab, @famosab)
 
 ### `Fixed`
 

conf/modules.config

@@ -198,4 +198,34 @@ process {
             mode: params.publish_dir_mode,
         ]
     }
+
+     withName: CPSR_RUN {
+        ext.prefix   = { "${meta.patient}.${meta.sample}" }
+        ext.genome   = { params.genome.toLowerCase() }
+        ext.database = { params.database }
+        ext.args     = {
+            [
+                "",
+                // CPSR params
+                "--panel_id ${params.panel_id}",
+                params.diagnostic_grade_only    ? "--diagnostic_grade_only" : "",
+                params.ignore_noncoding         ? "--ignore_noncoding" : "",
+                params.pop_gnomad               ? "--pop_gnomad ${params.pop_gnomad}" : "",
+                params.maf_upper_threshold      ? "--maf_upper_threshold ${params.maf_upper_threshold}" : "",
+                params.classify_all             ? "--classify_all" : "",
+                params.clinvar_ignore_noncancer ? "--clinvar_ignore_noncancer" : "",
+
+                // VEP params
+                params.vep_n_forks              ? "--vep_n_forks ${params.vep_n_forks}" : "",
+                params.vep_buffer_size          ? "--vep_buffer_size ${params.vep_buffer_size}" : "",
+                params.vep_gencode_basic        ? "--vep_gencode_basic" : "",
+                params.vep_pick_order           ? "--vep_pick_order ${params.vep_pick_order}" : "",
+                params.vep_no_intergenic        ? "--vep_no_intergenic" : "",
+            ].join(" ").trim()
+        }
+        publishDir = [
+            path: { "${params.outdir}/cpsr" },
+            mode: params.publish_dir_mode,
+        ]
+    }
 }

conf/test.config

@@ -43,4 +43,6 @@ params {
     filter_strelka_snvs        = "-i 'FILTER=\"PASS\"'"
     // html report generation throws errors with downsampled cache
     no_html                    = true
+    vep_gencode_basic          = true
+    vep_no_intergenic          = true
 }

docs/output.md

@@ -13,7 +13,8 @@ The pipeline is built using [Nextflow](https://www.nextflow.io/) and processes d
 - [Reference data](#reference-data) – PCGR and VEP resources when downloaded or provided as archives
 - [VCF preprocessing](#vcf-preprocessing) – optional bgzip/tabix, left-normalisation and filtering
 - [Variant formatting and intersection](#variant-formatting-and-intersection) – reheader VCFs, merge caller support, and reformat CNAs
-- [PCGR](#pcgr) – combined variant annotation and reporting
+- [PCGR](#pcgr) – combined somatic variant annotation and reporting
+- [CPSR](#cpsr) – combined germline variant annotation and reporting
 - [MultiQC](#multiqc) – aggregate QC and provenance reporting
 - [Pipeline information](#pipeline-information) – run metadata and software versions
 
@@ -80,6 +81,21 @@ Variants are first intersected across somatic callers to capture support per too
 
 PCGR ingests the unified somatic VCF (and optional allele-specific CNA table) together with the provided reference bundle and VEP cache to produce interactive and machine-readable reports per sample.
 
+### CPSR
+
+<details markdown="1">
+<summary>Output files</summary>
+
+- `cpsr/{sample}/`
+  - `{sample}.cpsr.{assembly}.html`: interactive CPSR report.
+  - `{sample}.cpsr.{assembly}.xlsx`: Excel summary of annotated variants.
+  - `{sample}.cpsr.{assembly}.pass.*`: PASS-filtered TSV/VCF summaries (e.g. `.pass.tsv.gz`, `.pass.vcf.gz` + `.tbi`).
+  - `{sample}.cpsr.{assembly}.conf.yaml`: configuration used by CPSR.
+
+</details>
+
+CPSR ingests each germline VCF together with the provided reference bundle and VEP cache to produce interactive and machine-readable reports per sample.
+
 ### MultiQC
 
 <details markdown="1">

docs/usage.md

@@ -6,7 +6,7 @@
 
 ## Introduction
 
-nf-core/variantprioritization consumes small variant calls (SNVs/InDels) in VCF format and optionally copy number alteration (CNA) segments, and produces annotated reports using PCGR.
+nf-core/variantprioritization consumes small variant calls (SNVs/InDels) in VCF format and optionally copy number alteration (CNA) segments, and produces annotated reports for the somatic case using PCGR / for the germline case using CPSR.
 
 The pipeline is designed to work well with outputs from nf-core/sarek (e.g. Mutect2, Strelka), but any compatible VCFs can be used.
 
@@ -22,7 +22,9 @@ It must be a comma-separated file with a header row and the following required c
 - `sample`: Sample identifier (no spaces)
 - `vcf`: Path to a VCF / VCF.GZ file
 
-Optionally, you can provide:
+The PCGR / CPSR implementation runs based on the input: If the samplesheet contains both status `0` and `1`, both reporting tools will be used.
+
+Optionally (somatic analysis only), you can provide:
 
 - `cna`: Path to a CNA segments file (required for somatic rows when `--cna_analysis` is enabled)
 
@@ -34,6 +36,9 @@ Optionally, you can provide:
 
 If you have multiple VCFs for the same sample (e.g. one per caller), provide one row per VCF. The pipeline will keep caller-specific files separate during preprocessing and then consolidate/intersect somatic calls per sample for PCGR reporting.
 
+> [!NOTE]
+> Intersection of calls is only implemented for the somatic reporting. For germline calls please provide unique sample ids.
+
 ### CNA files
 
 If you run with `--cna_analysis`, CNA segments are required for somatic (`status=1`) rows.

modules/local/cpsr/run/environment.yml

@@ -0,0 +1,7 @@
+channels:
+  - pcgr
+  - conda-forge
+  - bioconda
+dependencies:
+  - pcgr::pcgr=2.2.1
+  - pcgr::r-pcgrr=2.2.1

modules/local/cpsr/run/main.nf

@@ -0,0 +1,48 @@
+process CPSR_RUN {
+    tag "${meta.patient}.${meta.sample}"
+    label 'process_medium'
+
+    conda "${moduleDir}/environment.yml"
+    container "${workflow.containerEngine == 'singularity' && !task.ext.singularity_pull_docker_container
+        ? 'docker.io/sigven/pcgr:2.2.1'
+        : 'docker.io/sigven/pcgr:2.2.1'}"
+
+    input:
+    tuple val(meta), path(vcf), path(tbi)
+    path pcgr_dir
+    path vep_cache
+
+    output:
+    tuple val(meta), path("${prefix}"), emit: cpsr_reports
+    tuple val("${task.process}"), val('cpsr'),  eval("cpsr --version | sed 's/cpsr //g'"), topic: versions, emit: versions_cpsr
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    def genome   = task.ext.genome ?: ''
+    def args     = task.ext.args ?: ''
+    prefix   = task.ext.prefix ?: "${meta.id}"
+    """
+    export XDG_CACHE_HOME=/tmp
+    export XDG_DATA_HOME=/tmp
+    export QUARTO_PRINT_STACK=true
+
+    mkdir -p ${prefix}
+
+    cpsr \\
+        --input_vcf ${vcf} \\
+        --vep_dir ${vep_cache} \\
+        --refdata_dir ${pcgr_dir} \\
+        --output_dir ${prefix} \\
+        --genome_assembly ${genome} \\
+        --sample_id ${prefix} \\
+         ${args}
+    """
+
+    stub:
+    prefix = task.ext.prefix ?: "${meta.id}"
+    """
+    mkdir -p ${prefix}
+    """
+}

modules/local/pcgr/run/environment.yml

@@ -4,3 +4,4 @@ channels:
   - bioconda
 dependencies:
   - pcgr::pcgr=2.2.1
+  - pcgr::r-pcgrr=2.2.1

nextflow.config

@@ -150,9 +150,6 @@ params {
     outdir                          = null
     email                           = null
     multiqc_title                   = null
-    outdir                          = null
-    email                           = null
-    multiqc_title                   = null
 }
 
 // Load base.config by default for all pipelines

subworkflows/local/vcf_preprocessing/main.nf → subworkflows/local/file_preprocessing/main.nf

@@ -4,18 +4,18 @@
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 */
 
-include { TABIX_BGZIPTABIX } from '../../../modules/nf-core/tabix/bgziptabix'
-include { TABIX_TABIX      } from '../../../modules/nf-core/tabix/tabix'
-include { BCFTOOLS_NORM    } from '../../../modules/nf-core/bcftools/norm'
-include { BCFTOOLS_FILTER  } from '../../../modules/nf-core/bcftools/filter'
+include { BCFTOOLS_FILTER   } from '../../../modules/nf-core/bcftools/filter'
+include { BCFTOOLS_NORM     } from '../../../modules/nf-core/bcftools/norm'
+include { TABIX_BGZIPTABIX  } from '../../../modules/nf-core/tabix/bgziptabix'
+include { TABIX_TABIX       } from '../../../modules/nf-core/tabix/tabix'
 
 /*
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
     RUN MAIN WORKFLOW
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 */
 
-workflow VCF_PREPROCESSING {
+workflow FILE_PREPROCESSING {
     take:
     ch_samplesheet // channel: samplesheet read in from --input and processed by PIPELINE_INITIALISATION
     fasta
@@ -60,15 +60,21 @@ workflow VCF_PREPROCESSING {
 
     // Create index for files that need tabix
     ch_vcf.to_tabix
-        .map { meta, vcf, _tbi ->
-            vcf: [meta, vcf]
-        }
+        .map { meta, vcf, _tbi -> tuple(meta.id, meta, vcf) }
+        .set { ch_vcf_to_tabix_keyed }
+
+    ch_vcf_to_tabix_keyed
+        .map { _id, meta, vcf -> [meta, vcf] }
         .set { ch_vcf_to_tabix }
 
     TABIX_TABIX(ch_vcf_to_tabix)
 
-    ch_vcf_to_tabix
-        .join(TABIX_TABIX.out.index)
+    ch_vcf_to_tabix_keyed
+        .join(
+            TABIX_TABIX.out.index.map { meta, tbi -> tuple(meta.id, meta, tbi) },
+            by: 0
+        )
+        .map { _id, meta, vcf, _meta2, tbi -> [meta, vcf, tbi] }
         .set { ch_vcf_with_tabix }
 
     // Combine all processed files into a final channel
@@ -84,7 +90,11 @@ workflow VCF_PREPROCESSING {
     BCFTOOLS_FILTER(norm_ch)
     filtered_ch = BCFTOOLS_FILTER.out.vcf.join(BCFTOOLS_FILTER.out.tbi)
 
+    normalised_germline = filtered_ch.filter { meta, _vcf, _tbi -> meta.status == 'germline' }
+    normalised_somatic = filtered_ch.filter { meta, _vcf, _tbi -> meta.status == 'somatic' }
+
     emit:
-    filtered_ch // channel: [ meta, path(vcf_file), path(tbi_file), path(cna_file) ]
+    normalised_germline
+    normalised_somatic
     ch_cna_files
 }

subworkflows/local/format_files/main.nf → subworkflows/local/prepare_somatic/main.nf

@@ -15,7 +15,7 @@ include { PCGR_PREPAREVCF } from '../../../modules/local/pcgr/preparevcf'
 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
 */
 
-workflow FORMAT_FILES {
+workflow PREPARE_SOMATIC {
     take:
     vcf_files
     cna_files