main.nf

#!/usr/bin/env nextflow

nextflow.enable.dsl = 2

process filter_by_samplesheet{

    publishDir params.outdir, mode: 'copy'

    input:
    path(cgmlsttab)
    path(samplesheet)

    output:
    path("filtered_cgmlst.tab")

    """
    cat ${samplesheet} | cut -d',' -f1 > ids.csv
    awk -F '\t' 'NR==FNR {id[\$1]; next} \$1 in id' ids.csv ${cgmlsttab} > filtered_cgmlst.tab
    
    """
}


process calculate_distance {
    publishDir params.outdir, mode: 'copy'

    input:
    path(cgmlst_csv)

    output:
    path("distance.csv")

    script:
    if( params.mode == 'count-missing')
        """
        cat ${cgmlst_csv} | sed 's/,/\t/g' > cgmlst.tab
        cgmlst-dists-count-missing-as-diff -c cgmlst.tab > distance.csv
        """
    else
        """
        cat ${cgmlst_csv} | sed 's/,/\t/g' > cgmlst.tab
        cgmlst-dists -c cgmlst.tab > distance.csv
        """
}

process dendrogram {
    publishDir params.outdir, mode: 'copy'

    input:
    path(distance_matrix)

    output:
    path("dendrogram.pdf")

    script:
    """
    dendrogram.py ${distance_matrix}

    """

}

process cluster_py {
    publishDir params.outdir, mode: 'copy'
    
    when:
    params.runClustering == true

    input:
    path(distance_matrix)

    output:
    path("${params.linkage_type}_cluster_*.csv")

    script:
    """

    for i in ${params.threshold}; do cluster.py  ${distance_matrix} -t \${i} --linkage ${params.linkage_type} | awk 'BEGIN {OFS=","; print "run_accession","our_clusters"} {print \$0}' > ${params.linkage_type}_cluster_\${i}.csv; done
   
    """
}


workflow {

   
    ch_cgmlst = Channel.fromPath(params.cgmlst)

    if(params.samplesheet_input != 'NO_FILE'){
    //filter out combined_cgmlst.csv based on sample ids in the sample sheet
    ch_samplesheet = Channel.fromPath(params.samplesheet_input)   
    
    filter_by_samplesheet(ch_cgmlst, ch_samplesheet)    
    calculate_distance(filter_by_samplesheet.out)	
    }else{

    calculate_distance(ch_cgmlst)
    
   }
    dendrogram(calculate_distance.out)
    cluster_py(calculate_distance.out)


}