Have a functional version

Author: merszym

.gitignore

@@ -1,7 +1,5 @@
 .nextflow*
 work*/
-Downloads/
-Database/
-Kraken/
+out/
 conda/
 

main.nf

@@ -2,11 +2,43 @@
 
 params.kraken   =   "/home/merlin_szymanski/Kraken/install/"
 params.kmers    =   ["22"]
-params.outdir   =   "./out"
+params.outdir   =   false
+
+def helpMessage(){
+    log.info"""
+    DATASTRUCTURE PIPELINE
+
+    Download all mammalian mitochondiral genomes from the current NCBI/Refseq
+    release and create the datastructure required by the sediment_nf pipeline.
+    Creates 4 folders:
+    
+    1. ncbi:     raw Downloaded files from NCBI
+    2. genomes:  Fasta files grouped by Family
+    3. masked:   For all fasta files in 'genomes' a masked bed file
+    4. kraken:   For the given kmers the kraken databases
+    
+    USAGE:
+    nextflow run path/to/main.nf --outdir ABSOLUTE_PATH
+
+    required:
+        --outdir  PATH:   absolute path to the save-dir. e.g. "/mnt/scratch/.../out"
+
+    optional:
+        --kraken  PATH:   path to your kraken installation folder.
+                          default: '/home/merlin_szymanski/Kraken/install'
+        
+        --kmers   ARRAY:  Array of kmers for which databases should be created (this is not tested) 
+                          default: '["22"]'
+    """.stripIndent()
+}
+if(params.outdir == false){
+    helpMessage()
+    exit 0
+}
 
 
 process downloadGenomes{
-    publishDir "${params.outdir}/Downloads", mode: 'link'
+    publishDir "${params.outdir}/ncbi", mode: 'link'
     tag "Downloading..."
 
     output:
@@ -36,28 +68,45 @@ process extractFamilies{
 }
 
 extracted_fasta
-    .map{[it.baseName.split("_")[0], it.baseName.split("_")[1..2].join("_"), file(it)]}
+    .map{[it.baseName.split("_")[0], it.baseName.split("_")[1..-1].join("_"), file(it)]}
     .set{extracted_fasta}
 
 
 process writeFastas{
-    publishDir "${params.outdir}/Database/genomes/${family}/", saveAs: {"${species}.fasta"}
+    publishDir "${params.outdir}/genomes/${family}/", saveAs: {"${species}.fasta"}, pattern: "*.fasta", mode:'link'
+
     tag "$family:$species"
 
     input:
-        set family, species, file(fasta) from extracted_fasta
+        set family, species, "input.fasta" from extracted_fasta
 
     output:
-        set family, species, file(fasta) into (for_bed, for_kraken)
+        set family, species, "output.fasta" into (for_bed, for_bwa, for_kraken)
+
+    script:
+        """
+        cat input.fasta > output.fasta
+        """
+}
 
+process indexFasta{
+    publishDir "${params.outdir}/genomes/${family}/", mode: 'link'
+    tag "$family:$species"
+    
+    input:
+        set family, species, "${species}.fasta" from for_bwa
+
+    output:
+	file "${species}.fasta.*"
+   
     script:
         """
-        touch "$fasta"
+        bwa index "${species}.fasta"
         """
 }
 
 process writeBedFiles{
-    publishDir "${params.outdir}/Database/masked/${family}/", saveAs: {"${species}.masked.bed"}
+    publishDir "${params.outdir}/masked/", saveAs: {"${species}.masked.bed"}, mode:'link'
     tag "$family:$species"
 
     input:
@@ -71,7 +120,6 @@ process writeBedFiles{
         dustmasker -in species.fasta -outfmt acclist | \
         python3 $baseDir/bin/dustmasker_interval_to_bed.py \
         > species.masked.bed;
-
         """
 }
 
@@ -82,17 +130,14 @@ for_kraken
 
 process createKrakenDB{
     conda "$baseDir/envs/environment.yml"
-    publishDir "${params.outdir}/Kraken/${dbname}", pattern: "*.{tmp, idx, kdb, txt}"
-    publishDir "${params.outdir}/Kraken/${dbname}/taxonomy", pattern: "*.{dmp}"
     tag "THIS TAKES A LONG TIME"
 
     input:
         each kmer from params.kmers
         file fasta_list from for_kraken
 
     output:
-        file "${dbname}/taxonomy/*.dmp"
-	file "${dbname}/*.{tmp, idx, kdb, txt}"
+        file "output.txt" into log
 
     script:
         dbname = "Mito_db_kmer${kmer}"
@@ -104,7 +149,12 @@ process createKrakenDB{
 		done
         ${params.kraken}/kraken-build --build --db ${dbname} --kmer $kmer
         ${params.kraken}/kraken-build --clean --db ${dbname}
-        find ${dbname} -type f -exec touch {} +
+	if [[ -d ${params.outdir}/kraken ]];\
+		then rm -fr ${params.outdir}/kraken;\
+		fi;
+	mkdir ${params.outdir}/kraken
+	cp -r ${dbname} ${params.outdir}/kraken
+	touch "output.txt"
 	"""
 }