Allow Nextflow to manage Docker FS mounts

Former-commit-id: 3d787f8

modules/lineage.nf

@@ -6,27 +6,24 @@ process GET_POPPUNK_DB {
 
     input:
     val db_remote
-    val local
+    path local
 
     output:
-    env POPPUNK_DB_DIR
+    tuple path(local), env(DB_NAME)
 
     shell:
     '''
     DB_NAME=$(basename !{db_remote} .tar.gz)
 
     if [ ! -f !{local}/$DB_NAME/done_poppunk_$DB_NAME ] || [ ! -f !{local}/$DB_NAME/$DB_NAME.h5 ]; then
         rm -rf !{local}/$DB_NAME
-        mkdir -p !{local}
 
         wget !{db_remote} -O poppunk_db.tar.gz
         tar -xzf poppunk_db.tar.gz -C !{local}
         rm poppunk_db.tar.gz
 
         touch !{local}/$DB_NAME/done_poppunk_$DB_NAME
     fi
-
-    POPPUNK_DB_DIR=!{local}/$DB_NAME
     '''
 }
 
@@ -38,25 +35,22 @@ process GET_POPPUNK_EXT_CLUSTERS {
 
     input:
     val ext_clusters_remote
-    val local
+    path local
 
     output:
-    env EXT_CLUSTERS_PATH
+    env EXT_CLUSTERS_FILE
 
     shell:
     '''
     EXT_CLUSTERS_FILE=$(basename !{ext_clusters_remote})
 
     if [ ! -f !{local}/$EXT_CLUSTERS_FILE ] || [ ! -f !{local}/done_$EXT_CLUSTERS_FILE ]; then
-        rm -f !{local}/$EXT_CLUSTERS_FILE
-        mkdir -p !{local}
+        rm -f !{local}/$EXT_CLUSTERS_FILE !{local}/done_$EXT_CLUSTERS_FILE
 
         wget !{ext_clusters_remote} -O !{local}/$EXT_CLUSTERS_FILE
 
         touch !{local}/done_$EXT_CLUSTERS_FILE
     fi
-
-    EXT_CLUSTERS_PATH=!{local}/$EXT_CLUSTERS_FILE
     '''
 }
 
@@ -67,8 +61,8 @@ process LINEAGE {
     label 'poppunk_container'
 
     input:
-    path poppunk_db
-    path poppunk_ext_clusters
+    tuple path(poppunk_dir), val(db_name)
+    val ext_clusters_file
     path poppunk_qfile
 
     output:
@@ -77,7 +71,7 @@ process LINEAGE {
     shell:
     '''
     sed 's/^/prefix_/' !{poppunk_qfile} > safe_qfile.txt
-    poppunk_assign --db !{poppunk_db} --external-clustering !{poppunk_ext_clusters} --query safe_qfile.txt --output output --threads $(nproc)
+    poppunk_assign --db !{poppunk_dir}/!{db_name} --external-clustering !{poppunk_dir}/!{ext_clusters_file} --query safe_qfile.txt --output output --threads $(nproc)
     sed 's/^prefix_//' output/output_external_clusters.csv > result.csv
     '''
 }

modules/mapping.nf

@@ -6,20 +6,21 @@ process GET_REF_GENOME_BWA_DB_PREFIX {
 
     input:
     path reference
-    val local
+    path local
 
     output:
-    val "$local/ref"
+    tuple path(local), env(PREFIX)
 
     shell:
     '''
-    if [ ! -f !{local}/done_bwa_db_!{reference} ] || [ ! -f !{local}/ref.amb ] || [ ! -f !{local}/ref.ann ] || [ ! -f !{local}/ref.bwt ] || [ ! -f !{local}/ref.pac ] || [ ! -f !{local}/ref.sa ] ; then
-        rm -rf !{local}
-        mkdir -p !{local}
+    PREFIX=ref
+    
+    if [ ! -f !{local}/done_bwa_db_!{reference} ] || [ ! -f !{local}/$PREFIX.amb ] || [ ! -f !{local}/$PREFIX.ann ] || [ ! -f !{local}/$PREFIX.bwt ] || [ ! -f !{local}/$PREFIX.pac ] || [ ! -f !{local}/$PREFIX.sa ] ; then
+        rm -rf !{local}/{,.[!.],..?}*
 
-        bwa index -p ref !{reference}
+        bwa index -p $PREFIX !{reference}
         
-        mv ref.amb ref.ann ref.bwt ref.pac ref.sa -t !{local}
+        mv $PREFIX.amb $PREFIX.ann $PREFIX.bwt $PREFIX.pac $PREFIX.sa -t !{local}
         
         touch !{local}/done_bwa_db_!{reference}
     fi 
@@ -32,15 +33,15 @@ process MAPPING {
     label 'bwa_container'
 
     input:
-    val reference_prefix
+    tuple path(bwa_ref_db_dir), val(prefix)
     tuple val(sample_id), path(read1), path(read2), path(unpaired)
 
     output:
     tuple val(sample_id), path("${sample_id}_mapped.sam"), emit: sam
 
     shell:
     '''
-    bwa mem -t $(nproc) !{reference_prefix} <(zcat -f -- < !{read1}) <(zcat -f -- < !{read2}) > !{sample_id}_mapped.sam
+    bwa mem -t $(nproc) !{bwa_ref_db_dir}/!{prefix} <(zcat -f -- < !{read1}) <(zcat -f -- < !{read2}) > !{sample_id}_mapped.sam
     '''
 }
 

modules/serotype.nf

@@ -6,7 +6,7 @@ process GET_SEROBA_DB {
 
     input:
     val remote
-    val local
+    path local
 
     output:
     env CREATE_DB, emit: create_db
@@ -15,7 +15,7 @@ process GET_SEROBA_DB {
     '''
     # Assume up-to-date if done_seroba exists and the host cannot be resolved (often means the Internet is not available)
     if [ ! -f !{local}/done_seroba ] || !((git -C !{local} pull || echo 'Already up-to-date') | grep -q 'Already up[- ]to[- ]date'); then
-        rm -rf !{local}
+        rm -rf !{local}/{,.[!.],..?}*
         git clone !{remote} !{local}
 
         CREATE_DB=true
@@ -31,17 +31,19 @@ process CREATE_SEROBA_DB {
     label 'seroba_container'
 
     input:
-    val local
+    path seroba_dir
     val create_db
 
     output:
-    val "$local/database"
+    tuple path(seroba_dir), env(DATABASE)
 
     shell:
     '''
+    DATABASE=database
+
     if [ !{create_db} = true ]; then
-        seroba createDBs !{local}/database/ 71
-        touch !{local}/done_seroba
+        seroba createDBs !{seroba_dir}/$DATABASE/ 71
+        touch !{seroba_dir}/done_seroba
     fi
     '''
 }
@@ -51,15 +53,15 @@ process SEROTYPE {
     label 'seroba_container'
 
     input:
-    val database
+    tuple path(seroba_dir), val(database)
     tuple val(sample_id), path(read1), path(read2), path(unpaired)
 
     output:
     tuple val(sample_id), env(SEROTYPE), env(SEROBA_COMMENT), emit: result
 
     shell:
     '''
-    seroba runSerotyping !{database} !{read1} !{read2} !{sample_id}
+    seroba runSerotyping !{seroba_dir}/!{database} !{read1} !{read2} !{sample_id}
 
     SEROTYPE=$(awk -F'\t' '{ print $2 }' !{sample_id}/pred.tsv)
     SEROBA_COMMENT=$(awk -F'\t' '$3!=""{ print $3 } $3==""{ print "_" }' !{sample_id}/pred.tsv)

modules/taxonomy.nf

@@ -6,18 +6,17 @@ process GET_KRAKEN_DB {
 
     input:
     val remote
-    val local
+    path local
 
     output:
-    val local
+    path local
 
     shell:
     '''
     DB_NAME=$(basename !{remote})
 
     if [ ! -f !{local}/done_kraken_${DB_NAME} ] || [ ! -f !{local}/hash.k2d ]; then
-        rm -rf !{local}
-        mkdir -p !{local}
+        rm -rf !{local}/{,.[!.],..?}*
 
         wget !{remote} -O kraken_db.tar.gz
         tar -xzf kraken_db.tar.gz -C !{local}
@@ -33,7 +32,7 @@ process TAXONOMY {
     label 'kraken2_container'
 
     input:
-    val kraken_db
+    path kraken_db
     val kraken2_memory_mapping
     tuple val(sample_id), path(read1), path(read2), path(unpaired)
 

nextflow.config

@@ -98,5 +98,5 @@ process {
 // Enable the use of Docker container
 docker {
     enabled = true
-    runOptions = '-v $HOME:$HOME -u $(id -u):$(id -g)'
+    runOptions = '-u $(id -u):$(id -g)'
 }