Handle new reference naming

assets/report/report.qmd

@@ -293,12 +293,23 @@ if(params$run_ancestry == TRUE){
 ```{r colour_palette, echo = FALSE, eval=params$run_ancestry}
 # source: https://github.com/PGScatalog/PGS_Catalog/blob/master/catalog/static/catalog/pgs.scss#L2493-L2520
 # $ancestry_colours
-thousand_genomes_colours <- c("#FFD900", "#E41A1C", "#B15928", "#4DAF4A", 
-                              "#377EB8", "#00CED1", "#984EA3", "#A6CEE3", 
-                              "#FF7F00", "#BBB", "#999")
-names(thousand_genomes_colours) <- c("AFR", "AMR", "ASN", "EAS", "EUR", "GME",
-                                     "SAS", "MAE", "MAO", "NR", "OTH")
-thousand_genomes_palette <- scale_colour_manual(name = "Populations", values = thousand_genomes_colours)
+if({params$reference_panel_name} == '1000G'){
+  thousand_genomes_colours <- c("#FFD900", "#E41A1C", "#B15928", "#4DAF4A",
+                                "#377EB8", "#00CED1", "#984EA3", "#A6CEE3",
+                                "#FF7F00", "#BBB", "#999")
+  names(thousand_genomes_colours) <- c("AFR", "AMR", "ASN", "EAS",
+                                       "EUR", "GME", "SAS", "MAE",
+                                       "MAO", "NR", "OTH")
+  current_population_palette <- scale_colour_manual(name = "Populations", values = thousand_genomes_colours)
+} else if({params$reference_panel_name} == 'HGDP+1kGP'){
+  gnomAD_pop_colours <- c("#97519d", "#e42523", "#f67e1e", "#48b24b",
+                          "#3280bb", "#a65528", "#9a9c9b")
+  names(gnomAD_pop_colours) <- c("AFR", "AMR", "CSA", "EAS",
+                                 "EUR", "MID", "OCE")
+  current_population_palette <- scale_colour_manual(name = "Populations", values = gnomAD_pop_colours)
+} else{
+  current_population_palette <- scale_colour_brewer(palette="Set3")
+}
 ```
 
 ```{r, echo = FALSE, message = FALSE, eval=params$run_ancestry}
@@ -321,7 +332,7 @@ for(pc in seq.int(1,5,2)){
   if (pcX %in% colnames(popsim)){
     p_pca <- ggplot(popsim[popsim$REFERENCE == TRUE,], aes(x=!!sym(pcX), y=!!sym(pcY))) + geom_point(aes(colour=SuperPop, shape=slabel), alpha=0.25)
     p_pca <- p_pca + geom_point(data=popsim[popsim$REFERENCE != TRUE,], aes(color=MostSimilarPop, shape=slabel))
-    p_pca <- p_pca + theme_bw() + thousand_genomes_palette + scale_shape_manual(values=map_shapes, name='sampleset')
+    p_pca <- p_pca + theme_bw() + current_population_palette + scale_shape_manual(values=map_shapes, name='sampleset')
     print(p_pca)
   }
 }
@@ -492,4 +503,4 @@ For scores from the PGS Catalog, please remember to cite the original publicatio
 
 > PGS Catalog Calculator (in development). PGS Catalog Team. `https://github.com/PGScatalog/pgsc_calc`
 
-> Lambert et al. (2021) The Polygenic Score Catalog as an open database for reproducibility and systematic evaluation. Nature Genetics. 53:420–425 doi:10.1038/s41588-021-00783-5.
+> Lambert et al. (2021) The Polygenic Score Catalog as an open database for reproducibility and systematic evaluation. Nature Genetics. 53:420–425 doi:10.1038/s41588-021-00783-5.

conf/base.config

@@ -51,4 +51,7 @@ process {
     withName:DUMPSOFTWAREVERSIONS {
         cache = false
     }
+    withLabel:plink2{
+        memory = { check_max( 16.GB * task.attempt, 'memory' ) }
+    }
 }

modules/local/ancestry/bootstrap/make_database.nf

@@ -14,6 +14,8 @@ process MAKE_DATABASE {
 
     input:
     path '*'
+    tuple val(grch37_king_meta), path(grch37_king)
+    tuple val(grch38_king_meta), path(grch38_king)
     path checksums
 
     output:
@@ -26,6 +28,12 @@ process MAKE_DATABASE {
 
     echo $workflow.manifest.version > meta.txt
 
+    # can't use meta variables in stageAs
+    # don't want to use renameTo because it's destructive for the input
+    cp -L $grch37_king ${grch37_king_meta.build}_${grch37_king_meta.id}.king.cutoff.out.id
+    cp -L $grch38_king ${grch38_king_meta.build}_${grch38_king_meta.id}.king.cutoff.out.id
+    rm $grch37_king $grch38_king
+
     tar --dereference -acf pgsc_calc.tar.zst *
 
     cat <<-END_VERSIONS > versions.yml

modules/local/ancestry/extract_database.nf

@@ -17,18 +17,17 @@ process EXTRACT_DATABASE {
 
     output:
     tuple val(meta38), path("GRCh38_*_ALL.pgen"), path("GRCh38_*_ALL.psam"), path("GRCh38_*_ALL.pvar.zst"), emit: grch38, optional: true
-    tuple val(meta38), path("deg2_hg38.king.cutoff.out.id"), emit: grch38_king, optional: true
+    tuple val(meta38), path("GRCh38_*.king.cutoff.out.id"), emit: grch38_king, optional: true
     tuple val(meta37), path("GRCh37_*_ALL.pgen"), path("GRCh37_*_ALL.psam"), path("GRCh37_*_ALL.pvar.zst"), emit: grch37, optional: true
-    tuple val(meta37), path("deg2_phase3.king.cutoff.out.id"), emit: grch37_king, optional: true
+    tuple val(meta37), path("GRCh37_*.king.cutoff.out.id"), emit: grch37_king, optional: true
     path "versions.yml", emit: versions
 
     script:
     meta38 = ['build': 'GRCh38']
     meta37 = ['build': 'GRCh37']
-    def king = params.target_build == "GRCh37" ? "deg2_phase3.king.cutoff.out.id" : "deg2_hg38.king.cutoff.out.id"
 
     """
-    tar -xvf $reference --wildcards "${params.target_build}*" $king
+    tar -xvf $reference --wildcards "${params.target_build}*"
 
     cat <<-END_VERSIONS > versions.yml
     ${task.process.tokenize(':').last()}:

modules/local/plink2_relabelpvar.nf

@@ -6,8 +6,8 @@ process PLINK2_RELABELPVAR {
 
     tag "$meta.id chromosome $meta.chrom"
 
-    storeDir ( params.genotypes_cache ? "$params.genotypes_cache/${meta.id}/${params.target_build}/${meta.chrom}" :
-              "$workDir/genomes/${meta.id}/${params.target_build}/${meta.chrom}/")
+    storeDir ( params.genotypes_cache ? "$params.genotypes_cache/${meta.id}/${meta.build}/${meta.chrom}" :
+              "$workDir/genomes/${meta.id}/${meta.build}/${meta.chrom}/")
 
     conda "${task.ext.conda}"
 
@@ -21,9 +21,9 @@ process PLINK2_RELABELPVAR {
     tuple val(meta), path(geno), path(pheno), path(variants)
 
     output:
-    tuple val(meta), path("*.pgen"), emit: geno
-    tuple val(meta), path("*.zst") , emit: variants
-    tuple val(meta), path("*.psam"), emit: pheno
+    tuple val(meta), path("${meta.build}_*.pgen"), emit: geno
+    tuple val(meta), path("${meta.build}_*.pvar.zst") , emit: variants
+    tuple val(meta), path("${meta.build}_*.psam"), emit: pheno
     tuple val(meta), path("*.vmiss.gz"), emit: vmiss
     path "versions.yml"            , emit: versions
 
@@ -49,11 +49,11 @@ process PLINK2_RELABELPVAR {
         $set_ma_missing \\
         --pfile ${geno.baseName} $compressed \\
         --make-just-pvar zs \\
-        --out ${params.target_build}_${prefix}${meta.chrom}
+        --out ${meta.build}_${prefix}${meta.chrom}
 
     # cross platform (mac, linux) method of preserving symlinks
-    cp -a $geno ${params.target_build}_${prefix}${meta.chrom}.pgen
-    cp -a $pheno ${params.target_build}_${prefix}${meta.chrom}.psam
+    cp -a $geno ${meta.build}_${prefix}${meta.chrom}.pgen
+    cp -a $pheno ${meta.build}_${prefix}${meta.chrom}.psam
     gzip *.vmiss
 
     cat <<-END_VERSIONS > versions.yml

subworkflows/local/ancestry/bootstrap_ancestry.nf

@@ -2,7 +2,7 @@
 // Create a database containing reference data required for ancestry inference
 //
 include { SETUP_RESOURCE } from '../../../modules/local/ancestry/bootstrap/setup_resource'
-include { PLINK2_RELABELPVAR } from '../../../modules/local/plink2_relabelpvar'
+include { PLINK2_RELABELPVAR as BOOTSTRAP_RELABEL } from '../../../modules/local/plink2_relabelpvar'
 include { MAKE_DATABASE } from '../../../modules/local/ancestry/bootstrap/make_database'
 
 workflow BOOTSTRAP_ANCESTRY {
@@ -33,11 +33,11 @@ workflow BOOTSTRAP_ANCESTRY {
 
     SETUP_RESOURCE.out.plink.dump( tag: 'ref_setup' )
 
-    PLINK2_RELABELPVAR( SETUP_RESOURCE.out.plink )
-    ch_versions = ch_versions.mix(PLINK2_RELABELPVAR.out.versions.first())
+    BOOTSTRAP_RELABEL( SETUP_RESOURCE.out.plink )
+    ch_versions = ch_versions.mix(BOOTSTRAP_RELABEL.out.versions.first())
 
-    PLINK2_RELABELPVAR.out.geno
-        .concat(PLINK2_RELABELPVAR.out.pheno, PLINK2_RELABELPVAR.out.variants)
+    BOOTSTRAP_RELABEL.out.geno
+        .concat(BOOTSTRAP_RELABEL.out.pheno, BOOTSTRAP_RELABEL.out.variants)
         .dump(tag: 'ancestry_relabelled')
         .set { relabelled }
 
@@ -47,12 +47,14 @@ workflow BOOTSTRAP_ANCESTRY {
         .groupTuple(size: 3)
         .dump(tag: 'ancestry_relabelled_grouped')
         .map { drop_meta_keys(it).flatten() }
-        .set{ relabelled_flat }
+        .set{ relabelled_flat }     
 
-    ref.king
-        .map { drop_meta_keys(it) }
-    // dropping meta keys simplifies the join
-        .join( relabelled_flat )
+    ref.king.branch {
+        GRCh37: it[0].build == "GRCh37"
+        GRCh38: it[0].build == "GRCh38"
+    }.set { ch_king }
+
+    relabelled_flat
         .flatten()
         .filter(Path)
         .collect()
@@ -62,7 +64,7 @@ workflow BOOTSTRAP_ANCESTRY {
     Channel.fromPath(params.ancestry_checksums, checkIfExists: true)
         .set { ch_checksums }
 
-    MAKE_DATABASE( ch_raw_ref, ch_checksums )
+    MAKE_DATABASE( ch_raw_ref, ch_king.GRCh37, ch_king.GRCh38, ch_checksums )
     ch_versions = ch_versions.mix(MAKE_DATABASE.out.versions)
 
     emit:

tests/modules/plink2/relabelpvar/main.nf

@@ -7,7 +7,7 @@ include { PLINK2_RELABELPVAR } from '../../../../modules/local/plink2_relabelpva
 
 workflow testrelabelpvar {
     vcf = file('https://gitlab.ebi.ac.uk/nebfield/test-datasets/-/raw/master/pgsc_calc/cineca_synthetic_subset.vcf.gz')
-    def meta = [id: 'test', chrom: 22]
+    def meta = [id: 'test', 'build': 'GRCh37', chrom: 22]
 
     PLINK2_VCF(Channel.of([meta, vcf]))