Add pgs subworkflow

Dockerfile

@@ -79,6 +79,3 @@ ENV JAVA_TOOL_OPTIONS="-Djdk.lang.Process.launchMechanism=vfork"
 
 COPY files/bin/trace /usr/bin/.
 COPY files/bin/vcf2geno /usr/bin/.
-
-
-

README.md

@@ -276,7 +276,7 @@ params.minimac_window = 100000
 ### Build docker image locally
 
 ```
-docker build -t genepi/nf-imputationserver:latest .
+docker build -t genepi/imputation-docker:latest .
 ```
 
 ### Run testcases

modules/local/ancestry_estimation/execute_trace.nf

@@ -18,12 +18,14 @@ process EXECUTE_TRACE {
     batch_name = "batch_${samples.baseName}"
 
     """
-    # extract samples form vcf
-    bcftools view --samples-file ${samples} -Oz ${vcf_file} > ${batch_name}.vcf.gz
-    tabix ${batch_name}.vcf.gz
+    tabix ${vcf_file}
 
-    # convert to geno. TODO: check peopleIncludeFile option instead of bcftools.
-    vcf2geno --inVcf ${batch_name}.vcf.gz --rangeFile ${reference_range} --out ${batch_name}
+    # convert to geno
+    vcf2geno \
+        --inVcf ${vcf_file} \
+        --rangeFile ${reference_range} \
+        --peopleIncludeFile ${samples} \
+        --out ${batch_name}
 
     # write config file for trace
     echo "GENO_FILE ${reference_geno}" > trace.config
@@ -34,7 +36,7 @@ process EXECUTE_TRACE {
     echo "OUT_PREFIX ${batch_name}" >> trace.config
 
     # execute trace with config file
-    trace -p trace.config > trace.log
+    trace -p trace.config
     """
 
 }

modules/local/ancestry_estimation/prepare_trace.nf

@@ -2,8 +2,6 @@ import groovy.json.JsonOutput
 
 process PREPARE_TRACE {
 
-    publishDir params.output, mode: 'copy'
-
     input:
     path(vcf_files)
     path(reference_sites)

modules/local/ancestry_estimation/visualize_ancestry.nf

@@ -25,7 +25,7 @@ process VISUALIZE_ANCESTRY {
         ),
         intermediates_dir='\$PWD',
         knit_root_dir='\$PWD',
-        output_file='\$PWD/08-estimated-populations.html'
+        output_file='\$PWD/estimated-populations.html'
     )"
     """
 

modules/local/compression/compression_encryption_vcf.nf

@@ -6,7 +6,7 @@ process COMPRESSION_ENCRYPTION_VCF {
     tag "Merge Chromosome ${chr}"
 
     input:
-    tuple val(chr), path(imputed_vcf_data), path(imputed_info), path(imputed_meta_vcf_data)
+    tuple val(chr), val(start), val(end), path(imputed_vcf_data), path(imputed_info), path(imputed_meta_vcf_data)
 
     output:
     path("*.zip"), emit: encrypted_file

modules/local/imputation/minimac4.nf

@@ -7,7 +7,7 @@ process MINIMAC4 {
     path minimac_map
 
     output:
-    tuple val(chr), file("*.dose.vcf.gz"), file("*.info"), file("*.empiricalDose.vcf.gz"), emit: imputed_chunks
+    tuple val(chr), val(start), val(end), file("*.dose.vcf.gz"), file("*.info"), file("*.empiricalDose.vcf.gz"), emit: imputed_chunks
 
     script:
     def map = minimac_map ? '--referenceEstimates --map ' + minimac_map : ''

modules/local/pgs_calculation/calculate_chunks.nf

@@ -0,0 +1,30 @@
+process CALCULATE_CHUNKS {
+
+    tag "${vcf_file}"
+
+    input:
+        tuple val(chr), val(start), val(end), file(vcf_file),  file(info_file),  file(empirical_vcf_file)
+        path(scores)
+
+    output:
+        path "*.txt", emit: scores_chunks
+        path "*.info", emit: info_chunks
+
+    script:
+        name = "${vcf_file.baseName}_${chr}_${start}_${end}"
+
+    """
+
+    pgs-calc apply ${vcf_file} \
+        --ref ${scores.join(',')} \
+        --out ${name}.scores.txt \
+        --info ${name}.scores.info \
+        --start ${start} \
+        --end ${end} \
+        ${params.pgs.fix_strand_flips ? "--fix-strand-flips" : ""} \
+        --min-r2 ${params.pgs.min_r2} \
+        --no-ansi
+
+    """
+
+}

modules/local/pgs_calculation/create_html_report.nf

@@ -0,0 +1,34 @@
+process CREATE_HTML_REPORT {
+
+    publishDir params.output, mode: 'copy'
+
+    input:
+    path(merged_score)
+    path(merged_info)
+    path(scores_meta)
+    path(estimated_ancestry)
+
+    output:
+    path "*.html", emit: html_report
+    path "*.coverage.txt", emit: coverage_report
+
+    script:
+    samples = params.ancestry.enabled ? "--samples ${estimated_ancestry}" : ""
+
+    """
+    pgs-calc report \
+        --data ${merged_score} \
+        --info ${merged_info} \
+        --meta ${scores_meta} \
+        $samples \
+        --out scores.html
+
+    pgs-calc report \
+        --data ${merged_score} \
+        --info ${merged_info} \
+        --meta ${scores_meta} \
+        --template txt \
+        --out scores.coverage.txt
+    """
+
+}

modules/local/pgs_calculation/merge_chunks_infos.nf

@@ -0,0 +1,18 @@
+process MERGE_CHUNKS_INFOS {
+
+  publishDir params.output, mode: 'copy'
+
+  input:
+    path(report_chunks)
+
+  output:
+    path "*.info", emit: merged_info_files
+
+  """
+
+  pgs-calc merge-info ${report_chunks} \
+    --out ${params.project}.info
+
+  """
+
+}

modules/local/pgs_calculation/merge_chunks_scores.nf

@@ -0,0 +1,18 @@
+process MERGE_CHUNKS_SCORES {
+
+  publishDir params.output, mode: 'copy'
+
+  input:
+    path(score_chunks)
+
+  output:
+    path "*.txt", emit: merged_score_files
+
+  """
+
+  pgs-calc merge-score ${score_chunks} \
+    --out ${params.project}.scores.txt
+
+  """
+
+}

nextflow.config

@@ -1,4 +1,3 @@
-
 manifest {
     name                                  = 'imputationserver2'
     version                               = '2.0.0-beta1'
@@ -52,6 +51,17 @@ params {
         threshold: 0.75
     ]
 
+    pgs = [ 
+        enabled: false,
+        min_r2: 0,
+        fix_strand_flips: false
+    ]
+
+    pgscatalog = [
+        scores: "",
+        meta: ""
+    ]
+
     // header information
     r2Filter               = 0
     password               = null

workflows/ancestry_estimation.nf

@@ -41,4 +41,7 @@ workflow ANCESTRY_ESTIMATION {
         references.first{it.getExtension()=="samples"}
     )
 
+    emit: 
+    estimated_ancestry =  ESTIMATE_ANCESTRY.out.populations.collect()
+
 }

workflows/imputation.nf

@@ -28,5 +28,5 @@ workflow IMPUTATION {
     )
 
     emit: 
-    imputed_chunks = MINIMAC4.out.imputed_chunks.groupTuple()
+    imputed_chunks = MINIMAC4.out.imputed_chunks
 }

workflows/imputationserver.nf

@@ -79,7 +79,7 @@ workflow IMPUTATIONSERVER {
                 )
 
                 ENCRYPTION(
-                    IMPUTATION.out
+                    IMPUTATION.out.groupTuple()
                 )
             }
         }
@@ -88,6 +88,15 @@ workflow IMPUTATIONSERVER {
     if (params.ancestry.enabled){
         ANCESTRY_ESTIMATION()
     }
+
+    if (params.pgs.enabled) {
+
+        PGS_CALCULATION(
+            IMPUTATION.out,
+            params.ancestry.enabled ? ANCESTRY_ESTIMATION.out : Channel.empty()
+        )
+
+    }
 
 }
 

workflows/pgs_calculation.nf

@@ -0,0 +1,36 @@
+include { CALCULATE_CHUNKS } from '../modules/local/pgs_calculation/calculate_chunks'
+include { MERGE_CHUNKS_INFOS } from '../modules/local/pgs_calculation/merge_chunks_infos'
+include { MERGE_CHUNKS_SCORES } from '../modules/local/pgs_calculation/merge_chunks_scores'
+include { CREATE_HTML_REPORT } from '../modules/local/pgs_calculation/create_html_report'
+
+workflow PGS_CALCULATION {
+
+    take: 
+    imputed_chunks
+    estimated_ancestry
+
+
+    main:    
+    scores = Channel.fromPath(params.pgscatalog.scores, checkIfExists:true).collect()
+
+    CALCULATE_CHUNKS(
+        imputed_chunks,
+        scores
+    )
+
+    MERGE_CHUNKS_SCORES(
+        CALCULATE_CHUNKS.out.scores_chunks.collect()
+    )
+
+    MERGE_CHUNKS_INFOS(
+        CALCULATE_CHUNKS.out.info_chunks.collect()
+    )
+
+    CREATE_HTML_REPORT(
+        MERGE_CHUNKS_SCORES.out.collect(),
+        MERGE_CHUNKS_INFOS.out.collect(),
+        file(params.pgscatalog.meta, checkIfExists:true),
+        estimated_ancestry.collect().ifEmpty([])
+    )
+
+}