add basic host filter step

jts · Jul 14, 2020 · 6c68916 · 6c68916
1 parent 42d198f
commit 6c68916
Show file tree

Hide file tree

Showing 3 changed files with 21 additions and 3 deletions.
diff --git a/conf/base.config b/conf/base.config
@@ -7,8 +7,9 @@ params{
     basecalled_fastq = false
     fast5_pass = false
     sequencing_summary = false
-    ref = false
-    bed = false
+    ref = '/.mounts/labs/simpsonlab/projects/ncov/ref/artic-ncov2019/primer_schemes/nCoV-2019/V3/nCoV-2019.reference.fasta'
+    bed = '/.mounts/labs/gsiprojects/genomics/SARS-CoV-2/PHOCOV2/ncov2019ArticNf.manual.V3/nCoV-2019.bed'
+    human_ref = '/.mounts/labs/simpsonlab/data/references/GRCh38_no_alt_analysis_set.GCA_000001405.15.fna'
     profile = false
 
     // Repo to download your primer scheme from
@@ -21,7 +22,7 @@ params{
     scheme =  'nCoV-2019'
 
     // Scheme version
-    schemeVersion = 'V2'
+    schemeVersion = 'V3'
 
     // Run experimental medaka pipeline? Specify in the command using "--medaka"
     medaka = false

diff --git a/main.nf b/main.nf
@@ -10,6 +10,7 @@ include printHelp from './modules/help.nf'
 include {articNcovNanopore} from './workflows/articNcovNanopore.nf' 
 include {ncovIllumina} from './workflows/illuminaNcov.nf'
 include {ncovIlluminaCram} from './workflows/illuminaNcov.nf'
+include {getSampleId; prepareFastqPair; performHostFilter} from './modules/utils'
 
 if (params.help){
     printHelp()
@@ -71,7 +72,10 @@ if ( ! params.prefix ) {
 
 // main workflow
 workflow {
+
    if ( params.illumina ) {
+       performHostFilter(prepareFastqPair.out)
+
        if (params.cram) {
            Channel.fromPath( "${params.directory}/**.cram" )
                   .map { file -> tuple(file.baseName, file) }

diff --git a/modules/utils.nf b/modules/utils.nf
@@ -0,0 +1,13 @@
+process performHostFilter {
+    input:
+        tuple(val(sampleId), path(forward), path(reverse))
+    output:
+        tuple sampleId, path("${sampleId}_hostfiltered_R1.fastq.gz"), path("${sampleId}_hostfiltered_R2.fastq.gz"), emit: fastqPairs
+
+    script:
+        """
+        bwa mem -t ${task.cpus} ${params.human_ref} ${forward} ${reverse} | samtools view -q 30 -U ${sampleId}.host_unaligned.bam -Sb > ${sampleId}.host_aligned.bam
+        samtools sort -n ${sampleId}.host_unaligned.bam | \
+             samtools fastq -1 ${sampleId}_hostfiltered_R1.fastq.gz -2 ${sampleId}_hostfiltered_R2.fastq.gz -s ${sampleId}_singletons.fastq.gz -
+        """
+}