Merge pull request #66 from hoelzer/fix/concat-carefully

Avoid zcat'ing *.gz when output file is also a .gz file.
rki-mf1 · Nov 15, 2023 · e87a382 · e87a382
2 parents 91865cb + 307ea83
commit e87a382
Showing 1 changed file with 26 additions and 25 deletions.
diff --git a/modules/prepare_contamination.nf b/modules/prepare_contamination.nf
@@ -16,30 +16,31 @@ process download_host {
 
   script:
   """
-  if [ $host == 'hsa' ]; then
-    wget ftp://ftp.ensembl.org/pub/release-99/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz
-    zcat *.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
-  fi
-  if [ $host == 'mmu' ]; then
-    wget ftp://ftp.ensembl.org/pub/release-99/fasta/mus_musculus/dna/Mus_musculus.GRCm38.dna.primary_assembly.fa.gz
-    zcat *.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
-  fi
-  if [ $host == 'cli' ]; then
-    wget ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/337/935/GCF_000337935.1_Cliv_1.0/GCF_000337935.1_Cliv_1.0_genomic.fna.gz
-    zcat *.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
-  fi
-  if [ $host == 'csa' ]; then
-    wget ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/409/795/GCF_000409795.2_Chlorocebus_sabeus_1.1/GCF_000409795.2_Chlorocebus_sabeus_1.1_genomic.fna.gz
-    zcat *.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
-  fi
-  if [ $host == 'gga' ]; then
-    wget ftp://ftp.ensembl.org/pub/release-99/fasta/gallus_gallus/dna/Gallus_gallus.GRCg6a.dna.toplevel.fa.gz
-    zcat *.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
-  fi
-  if [ $host == 'eco' ]; then
-    wget ftp://ftp.ensemblgenomes.org/pub/release-45/bacteria//fasta/bacteria_90_collection/escherichia_coli_k_12/dna/Escherichia_coli_k_12.ASM80076v1.dna.toplevel.fa.gz
-    zcat *.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
-  fi
+  case $host in
+    hsa)
+      wget ftp://ftp.ensembl.org/pub/release-99/fasta/homo_sapiens/dna/Homo_sapiens.GRCh38.dna.primary_assembly.fa.gz -O host-temp.fa.gz
+      ;;
+    mmu)
+      wget ftp://ftp.ensembl.org/pub/release-99/fasta/mus_musculus/dna/Mus_musculus.GRCm38.dna.primary_assembly.fa.gz -O host-temp.fa.gz
+      ;;
+    cli)
+      wget ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/337/935/GCF_000337935.1_Cliv_1.0/GCF_000337935.1_Cliv_1.0_genomic.fna.gz -O host-temp.fa.gz
+      ;;
+    csa)
+      wget ftp://ftp.ncbi.nlm.nih.gov/genomes/all/GCF/000/409/795/GCF_000409795.2_Chlorocebus_sabeus_1.1/GCF_000409795.2_Chlorocebus_sabeus_1.1_genomic.fna.gz -O host-temp.fa.gz
+      ;;
+    gga)
+      wget ftp://ftp.ensembl.org/pub/release-99/fasta/gallus_gallus/dna/Gallus_gallus.GRCg6a.dna.toplevel.fa.gz -O host-temp.fa.gz
+      ;;
+    eco)
+      wget ftp://ftp.ensemblgenomes.org/pub/release-45/bacteria//fasta/bacteria_90_collection/escherichia_coli_k_12/dna/Escherichia_coli_k_12.ASM80076v1.dna.toplevel.fa.gz -O host-temp.fa.gz
+      ;;
+    *)
+      echo "Unknown host ($host)."
+      ;;
+  esac
+
+  zcat host-temp.fa.gz | bgzip -@ ${task.cpus} -c > ${host}.fa.gz
   """
   stub:
   """
@@ -108,4 +109,4 @@ process concat_contamination {
   """
   touch db.fa.gz db.fa.fai
   """
-}
+}