From 7efb7d380e43643872c5738fd50ccffe6bd13235 Mon Sep 17 00:00:00 2001
From: Julian Thomas Mohr <julian.mohr@tron-mainz.de>
Date: Wed, 26 Jul 2023 13:42:40 +0200
Subject: [PATCH] Replace MERGE_REPLICATES by nf-core/module SAMTOOLS_MERGE and
 integrate into pipeline

---
 local_modules/singleton_input/main.nf | 17 ++++++
 main.nf                               | 82 ++++++++++++++++++++++++---
 2 files changed, 92 insertions(+), 7 deletions(-)
 create mode 100644 local_modules/singleton_input/main.nf

diff --git a/local_modules/singleton_input/main.nf b/local_modules/singleton_input/main.nf
new file mode 100644
index 0000000..a86047d
--- /dev/null
+++ b/local_modules/singleton_input/main.nf
@@ -0,0 +1,17 @@
+process SINGLETON_INPUT {
+    tag "$meta.id"
+
+    input:
+    tuple val(meta), path(input_bam)
+
+    output:
+    tuple val(meta), path("${meta.id}.bam")
+
+    when:
+    task.ext.when == null || task.ext.when
+
+    script:
+    """
+    ln -s $input_bam ${meta.id}.bam
+    """
+}
diff --git a/main.nf b/main.nf
index 4ee30e6..a1c9376 100755
--- a/main.nf
+++ b/main.nf
@@ -3,10 +3,12 @@
 
 nextflow.enable.dsl = 2
 
+include { SINGLETON_INPUT } from './local_modules/singleton_input/main'
+include { SAMTOOLS_MERGE } from './nf-core-modules/modules/nf-core/samtools/merge/main'
 include { CNVKIT_BATCH } from './nf-core-modules/modules/nf-core/cnvkit/batch/main'
 include { SEQUENZAUTILS_GCWIGGLE } from './nf-core-modules/modules/nf-core/sequenzautils/gcwiggle/main'
 include { SEQUENZAUTILS_BAM2SEQZ } from './nf-core-modules/modules/nf-core/sequenzautils/bam2seqz/main'
-include { MERGE_REPLICATES } from './local_modules/merge_replicates'
+//include { MERGE_REPLICATES } from './local_modules/merge_replicates'
 include { SEQUENZAUTILS_SEQZBINNING } from './local_modules/sequenzautils/seqzbinning/main'
 include { R_SEQUENZA } from './local_modules/rsequenza/main'
 
@@ -43,25 +45,91 @@ if (!params.input_files) {
 else {
   Channel
     .fromPath(params.input_files)
-    .splitCsv(header: ['name', 'tumor_bam', 'normal_bam'], sep: "\t")
-    .map{ row-> tuple([id: row.name], row.tumor_bam, row.normal_bam) }
+    .splitCsv(header: ['sample', 'tumor_bam', 'normal_bam'], sep: "\t")
+    .map{ row-> tuple([id: row.sample], row.tumor_bam, row.normal_bam) }
     .set { input_files }
 }
 
 workflow {
-    MERGE_REPLICATES(input_files)
-    merged_bams = MERGE_REPLICATES.out.merged_bams
+
+    tumor_bams = input_files
+        .map { 
+            meta, tumor_bam, normal_bam -> 
+            def fmeta = [:]
+            fmeta.id = meta.id + ".tumor"
+            fmeta.type = "tumor"
+            [fmeta, tumor_bam.tokenize(',')] 
+        }
+        .branch{
+            single: it[1].size() == 1
+            multiple: it[1].size() > 1
+        }
+
+    normal_bams = input_files
+        .map { 
+            meta, tumor_bam, normal_bam -> 
+            def fmeta = [:]
+            fmeta.id = meta.id + ".normal"
+            fmeta.type = "normal"
+            [fmeta, normal_bam.tokenize(',')] 
+        }
+        .branch{
+            single: it[1].size() == 1
+            multiple: it[1].size() > 1
+        }
+
+    SINGLETON_INPUT(tumor_bams.single.mix(normal_bams.single))
+
+    SAMTOOLS_MERGE(
+        tumor_bams.multiple.mix(normal_bams.multiple),
+        [[], []],
+        [[], []]
+    )
+
+    prepared_tumor_bams = SINGLETON_INPUT.out
+        .filter { it[0].type == "tumor" }
+        .mix(
+            SAMTOOLS_MERGE.out.bam
+                .filter { it[0].type == "tumor" }
+        ).map {
+            meta, bam ->
+            def fmeta = [:]
+            fmeta.id = meta.id[0..-(meta.type.length() + 2)]
+            [fmeta, bam]
+        }
+
+    prepared_normal_bams = SINGLETON_INPUT.out
+        .filter { it[0].type == "normal" }
+        .mix(
+            SAMTOOLS_MERGE.out.bam
+                .filter { it[0].type == "normal" }
+        ).map {
+            meta, bam ->
+            def fmeta = [:]
+            fmeta.id = meta.id[0..-(meta.type.length() + 2)]
+            [fmeta, bam]
+        }
+
+    ch_meta_tumor_normal = prepared_tumor_bams
+        .join(prepared_normal_bams, by: [0])
 
     if (params.toolslist.contains('cnvkit')) {
         // NOTE: it does not provide fasta.fai or CNVkit reference, but these are created every time
-        CNVKIT_BATCH(merged_bams, params.reference, [], params.intervals, [], false)
+        CNVKIT_BATCH(
+            ch_meta_tumor_normal,
+            params.reference, 
+            [], 
+            params.intervals, 
+            [], 
+            false
+        )
     }
 
     if (params.toolslist.contains('sequenza')) {
         SEQUENZAUTILS_GCWIGGLE([[id:'reference'], params.reference])
         wig = SEQUENZAUTILS_GCWIGGLE.out.wig.map { it[1] }
 
-        SEQUENZAUTILS_BAM2SEQZ(merged_bams, params.reference, wig)
+        SEQUENZAUTILS_BAM2SEQZ(ch_meta_tumor_normal, params.reference, wig)
 
         SEQUENZAUTILS_SEQZBINNING(SEQUENZAUTILS_BAM2SEQZ.out.seqz)