Observe the results
Does Snakemake do the job?
Fastqc: job or jobs?
-Look at more precisely the fastqc job. We have many input files but snakemake launched only one fastqc job:
It is because thefastqcrule is defined with a list of files and not for one unique file and because thefastqc tool accepts both a unique file as wellas a list of files.
Create snakemake file exo6.smk. Improve previous code: Thank to the all rule, all expected files are designated. So we don’t need to give the fastqc rule a list anymore and we can replace it to manage only one file and all files one by one. We will gain in power in system shaving more than one core.
HINT
+Look at more precisely the fastqc job. We have many input files but snakemake launched only one fastqc job:
It is because thefastqcrule is defined with a list of files and not for one unique file and because thefastqc tool accepts both a unique file as wellas a list of files.
Create snakemake file exo6.smk. Improve previous code: Thank to the all rule, all expected files are designated. So we don’t need to give the fastqc rule a list anymore and we can replace it to manage only one file and all files one by one. We will gain in power in system shaving more than one core.
HINT
Replace the expand()function with a wildcard for one filename in the fastqc rule.
Solution
SAMPLES = ["SRR3099585_chr18","SRR3099586_chr18","SRR3099587_chr18"]
BIDX = ["1","2","3","4","rev.1","rev.2"]
@@ -178,7 +178,7 @@
Snakemake run
rm -Rf FastQC/
snakemake -s exo6.smk -R all fastqc
Observe the results
-Now Snakemake did many fastqc jobs:
But what happens to the runtime displays on the screen? To correct this, we will move the displays to a log file specific for each rule and each input file.
Create snakemake file exo7.smk. Improve previous code:
+Now Snakemake did many fastqc jobs:
But what happens to the runtime displays on the screen? To correct this, we will move the displays to a log file specific for each rule and each input file.
Create snakemake file exo7.smk. Improve previous code:
In Unix systems, the output of a command is usually sent to two separate streams: the normal output: to Standard Out (stdout also “>” in shell),and error messages: to Standard Error (stderr, or “2>” in shell). To integrate stderr into the same log file as the stdout can be use “&>” instead of “>”:
shell: … &> {log}, but use with care when output files are printed tostdout (as often in shell comands). Redirect the stdout and stderr streams of the fastqc and bowtie2-build commands.
HINT
For the bowtie2-build and fastqc rules, add the log: directive with two variables (log1 and log2) to redirect each streams.
Solution
SAMPLES = ["SRR3099585_chr18","SRR3099586_chr18","SRR3099587_chr18"]
diff --git a/docs/assets/md/session2020/workflow.md b/docs/assets/md/session2020/workflow.md
index 03a3b9e..86406e9 100644
--- a/docs/assets/md/session2020/workflow.md
+++ b/docs/assets/md/session2020/workflow.md
@@ -242,7 +242,7 @@ Does Snakemake do the job?
***Fastqc: job or jobs?***
Look at more precisely the fastqc job. We have many input files but snakemake launched only one fastqc job:
-
+
It is because thefastqcrule is defined with a list of files and not for one unique file and because thefastqc tool accepts both a unique file as wellas a list of files.
@@ -294,7 +294,7 @@ snakemake -s exo6.smk -R all fastqc
***Observe the results***
Now Snakemake did many fastqc jobs:
-
+
But what happens to the runtime displays on the screen? To correct this, we will move the displays to a log file specific for each rule and each input file.
diff --git a/docs/session2020.html b/docs/session2020.html
index 550d574..6cd1ad1 100644
--- a/docs/session2020.html
+++ b/docs/session2020.html
@@ -403,7 +403,7 @@ Matériels des participants
Outils collaboratif pour le md : HackMD. Merci Clémence et Diane !