Add new de novo prediction features for wastewater and mixed samples

configs/container.config

@@ -20,4 +20,5 @@ process {
     withLabel:  snp_sites   { container = 'biocontainers/snp-sites:v2.4.1-1-deb_cv1' }
     withLabel:  toytree     { container = 'nanozoo/toytree:1.1.2--1295ae6' }
     withLabel:  ubuntu      { container = 'nanozoo/template:3.8--e13dfeb' }
+    withLabel:  devider     { container = 'nanozoo/devider:0.0.1--f255b80' }
 }

configs/local.config

@@ -22,4 +22,5 @@ process {
     withLabel:  snp_sites   { cpus = params.cores }
     withLabel:  seqrs       { cpus = 1 }
     withLabel:  ubuntu      { cpus = 1 } 
+    withLabel:  devider     { cpus = params.cores }
 }

configs/nodes.config

@@ -21,4 +21,5 @@ process {
     withLabel:  seqrs       { cpus = 2; memory = '2 GB' }
     withLabel:  snp_sites   { cpus = 4; memory = '4 GB' }
     withLabel:  ubuntu      { cpus = 2; memory = '2 GB' }
+    withLabel:  devider     { cpus = 12; memory = '12 GB'}
 }

poreCov.nf

@@ -321,6 +321,7 @@ include { genome_quality_wf } from './workflows/genome_quality.nf'
 include { read_classification_wf; read_screening_freyja_wf; read_screening_lsc_wf} from './workflows/read_classification'
 include { read_qc_wf } from './workflows/read_qc.nf'
 include { rki_report_wf } from './workflows/provide_rki.nf'
+include { ww_cryptic_lineages_wf } from './workflows/ww_cryptic_lineages.nf'
 
 /************************** 
 * MAIN WORKFLOW
@@ -419,6 +420,12 @@ workflow {
         determine_mutations_wf(fasta_input_ch)
         genome_quality_wf(fasta_input_ch, reference_for_qc_input_ch)
 
+    // TEST - devider haplotyping + freyja covariants co-occurring mutation profiles
+    // for cryptic variant detection later
+        if (!(params.fasta || workflow.profile.contains('test_fasta'))) { // create a flag/param for this? + make sure the channels exist?
+            ww_cryptic_lineages_wf(filtered_reads_ch, reference_for_qc_input_ch, artic_ncov_wf.out.trimmed_bam)
+        }
+
     // 3. Specialised outputs (rki, json)
         rki_report_wf(genome_quality_wf.out[0], genome_quality_wf.out[1], extended_input_ch)
 

workflows/process/devider.nf

@@ -0,0 +1,55 @@
+process devider {
+        label 'devider'
+        publishDir "${params.output}/${params.lineagedir}/devider", mode: 'copy', pattern: "*_majority_vote_haplotypes.fasta" // should this be published or only intermediate for later?
+        // probably change output dir
+
+    input:
+        tuple val(name), path(reads), path(reference)//, path(bam_file), path(vcf)
+    output:
+        tuple val(name), path("${name}_majority_vote_haplotypes.fasta"), emit: haplotypes
+        // add other stuff if relevant
+
+    script:
+    // just run devider haplotyping
+    /*"""
+    devider -b ${bam_file} -v ${vcf} -r ${reference} -o ${name}_devider_output \
+        -t ${task.cpus} --preset nanopore-r9
+    """*/ // didn't work when I tried it separately (because it doesnt recognize chromosome name or smth)
+
+    // run devider pipeline (does mapping, variant calling, haplotyping)
+    """
+    run_devider_pipeline -i ${reads} -r ${reference} -o ${name}_devider_output
+    mv ${name}_devider_output/majority_vote_haplotypes.fasta ${name}_majority_vote_haplotypes.fasta
+    """ 
+    /* different presets - choose one or make dynamic??
+        old-long-reads for ~90% identity long reads
+        nanopore-r9 for ~95% (default)
+        nanopore-r10 for ~98%
+        hi-fi for true high fidelity reads (> 99.9% accuracy).
+    */
+
+}
+
+
+process freyja_covariants {
+        label 'freyja'
+        publishDir "${params.output}/${params.lineagedir}/devider", mode: 'copy', pattern: "*"
+        // probably change output dir
+        // maybe move to freyja.nf for clarity?
+
+    input:
+        tuple val(name), path(bam_file), path(bam_index), path(reference)
+        // needs .bai in same folder as .bam
+
+    output:
+        tuple val(name), path("${name}_freyja_covariants.tsv"), emit: covariants
+
+    script:
+    """
+    freyja covariants --ref-genome ${reference} --output ${name}_freyja_covariants.tsv  ${bam_file} 
+    """
+    // could include .gff with --annot --> AA notes in result + nt mutation
+    //  --threads ${task.cpus} gets error "no such option as --threads" but it exists according to --help and the wiki ??
+    // potential issue: gives different result than running freyja covariants separately --> just because of different (trimmed) bam?
+
+}

workflows/ww_cryptic_lineages.nf

@@ -0,0 +1,18 @@
+include { devider; freyja_covariants } from './process/devider.nf'
+// adjust if freyja_covariants gets moved to freyja.nf
+
+workflow ww_cryptic_lineages_wf {
+    take:
+        fastq
+        reference
+        bam_and_bai
+
+    main:
+        devider(fastq.combine(reference)) 
+        freyja_covariants(bam_and_bai.combine(reference))
+
+    emit:
+        devider.out
+        freyja_covariants.out
+
+}