add fastqc rule

Snakefile

@@ -83,6 +83,8 @@ def makeDirectory(path_to_directory):
 
 
 def mkdir_results(results_subfolders):
+    if not config['fastqc']:
+        del results_subfolders['fastq_quality_control']
     for k in results_subfolders:
         subfolder=results_subfolders[k]
         if not os.path.exists(os.path.join("results",subfolder)):
@@ -150,7 +152,7 @@ def str_join(df, sep, *cols):
 # MAIN
 ###############################################################################
 
-
+## generate results subfolders
 mkdir_results(results_subfolders)
 
 ## check format (CLASSIC or RAPIDRUN)
@@ -170,8 +172,8 @@ if config['format'] == "CLASSIC":
             }
             dfrClassic = dfrClassic.append(thisRow, ignore_index=True)        
     print(dfrClassic)
-    rapidrunfile=results_subfolders['settings']+'/all_samples_classic.csv'
-    export_allsample = dfrClassic.to_csv (r'./'+rapidrunfile, index = None, header = False, sep = ';')
+    rapidrunfile='results/'+results_subfolders['settings']+'/all_samples_classic.csv'
+    export_allsample = dfrClassic.to_csv (r'{}'.format(rapidrunfile), index = None, header = False, sep = ';')
 else:
     print("RAPIDRUN data: many markers for many runs")
     #configfile: "01_infos/config.yaml"
@@ -249,7 +251,7 @@ for run in uniqRuns:
             dfMulti = dfMulti.append( thisRow, ignore_index=True)
 
 demultiplexFile='results/'+results_subfolders['settings']+'/all_demultiplex.csv'
-export_csv = dfMulti.to_csv (r'./'+demultiplexFile, index = None, header=True,sep=",")
+export_csv = dfMulti.to_csv (r'{}'.format(demultiplexFile), index = None, header=True,sep=",")
 
 print (dfMulti)
 
@@ -285,10 +287,10 @@ uniqRuns=dfrm.run.unique()
 
 ## projmarkrun wildcards
 #dfMultiProjMarkRunSample=dfMulti[['projet','marker','run','demultiplex']]
-dfMultiProjMarkRunSample=dfMulti.loc[:, ('projet','marker','run','demultiplex')]
+dfMultiProjMarkRunSample = dfMulti.loc[:, ('projet','marker','run','demultiplex')]
 dfMultiProjMarkRunSample["demultiplex"]="results/"+results_subfolders['demultiplex_tag']+"/samples/"+dfMultiProjMarkRunSample["demultiplex"]+".fastq"
 #dfMultiProjMarkRun=dfMulti[['projet','marker','run']]
-dfMultiProjMarkRun=dfMulti.loc[:, ('projet','marker','run')]
+dfMultiProjMarkRun = dfMulti.loc[:, ('projet','marker','run')]
 dfMultiProjMarkRun['projMarkRun']=str_join(dfMultiProjMarkRun, '/', 'projet', 'marker', 'run')
 dfMultiProjMarkRunSample['projMarkRun']=str_join(dfMultiProjMarkRunSample, '/', 'projet', 'marker', 'run')
 dfMultiProjMarkRun = dfMultiProjMarkRun.drop_duplicates()
@@ -296,18 +298,29 @@ print(dfMultiProjMarkRun)
 
 # projmark wildcards
 #dfMultiProjMark=dfMulti[['projet','marker']]
-dfMultiProjMark=dfMulti.loc[:, ('projet','marker')]
-dfMultiProjMark['projMark']=str_join(dfMultiProjMark, '/', 'projet', 'marker')
+dfMultiProjMark = dfMulti.loc[:, ('projet','marker')]
+dfMultiProjMark['projMark'] = str_join(dfMultiProjMark, '/', 'projet', 'marker')
 dfMultiProjMark = dfMultiProjMark.drop_duplicates()
 
 
 
+ruleAllInputList = [
+    'results/'+results_subfolders['flags']+'/fastq_quality_control.flag',
+    'results/'+results_subfolders['flags']+'/table_assigned_sequences.flag'
+    ]
+
+if not config['fastqc']:
+    ruleAllInputList.remove('results/'+results_subfolders['flags']+'/fastq_quality_control.flag')
+
 rule all:
     input:
-        'results/'+results_subfolders['flags']+'/table_assigned_sequences.flag'
-        
+        ruleAllInputList
 
-include: "rules/merge_fastq.smk"
+if config['fastqc']:
+    include: "rules/fastq_quality_control.smk"
+    include: "rules/merge_fastq_after_fastqc.smk"
+else:
+    include: "rules/merge_fastq.smk"
 include: "rules/demultiplex_tag.smk"
 
 

rules/fastq_quality_control.smk

+__author__ = "Pierre-Edouard Guerin"
+__license__ = "MIT"
+
+## Pool sequences and quality files
+rule fastq_quality_control:
+    input:
+        config["fichiers"]["folder_fastq"]+'{runR}.fastq.gz'
+    output:
+        'results/'+results_subfolders['fastq_quality_control']+'/{runR}_fastqc.zip'
+    params:
+        resFolder='results/'+results_subfolders['fastq_quality_control'],
+    conda:
+        '../envs/env_fastqc.yaml'
+    resources:
+        job=1
+    log:
+        'logs/'+results_subfolders['fastq_quality_control']+'/{runR}.log'
+    threads:
+        1
+    shell:
+        '''
+        fastqc --outdir {params.resFolder} {input} 2> {log}
+        '''
+
+
+rule checkpoint_fastqc:
+    input:
+        expand('results/'+results_subfolders['fastq_quality_control']+'/{runR}_fastqc.zip', runR=[str(r)+"_R2" for r in uniqRuns]+[str(r)+"_R1" for r in uniqRuns])
+    output:
+        touch('results/'+results_subfolders['flags']+'/fastq_quality_control.flag')
+

rules/merge_fastq_after_fastqc.smk

+__author__ = "Pierre-Edouard Guerin"
+__license__ = "MIT"
+
+
+## Merge read pairs but we have to trick fastqc output first
+rule merge_fastq_after_fastqc:
+    input:
+        'results/'+results_subfolders['flags']+'/fastq_quality_control.flag'       
+    output:
+        fq='results/'+results_subfolders['merge_fastq']+'/{run}.fastq'
+    singularity:
+        config["singularity"]["ednatools"]
+    conda:
+        '../envs/env_vsearch.yaml'
+    threads: 
+        config["merging"]["vsearch"]["cores"]
+    resources:
+        job=1
+    log:
+        'logs/'+results_subfolders['merge_fastq']+'/{run}.log'
+    params:
+        R1=config["fichiers"]["folder_fastq"]+'{run}_R1.fastq.gz',
+        R2=config["fichiers"]["folder_fastq"]+'{run}_R2.fastq.gz',
+        encoding=config["merge_fastq"]["encoding"]
+    shell:
+        '''ls {input};
+        vsearch \
+ --threads {threads} \
+ --fastq_mergepairs {params.R1} \
+ --reverse {params.R2} \
+ --fastq_ascii {params.encoding} \
+ --fastqout {output} \
+ --fastq_allowmergestagger \
+ --quiet 2> {log}'''
\ No newline at end of file