From da94c9ce1c200ea42893ec804206a291d8165401 Mon Sep 17 00:00:00 2001
From: Blaise Li <blaise.li@normalesup.org>
Date: Tue, 13 Dec 2016 14:12:40 +0100
Subject: [PATCH] Added scripts to run and parse fastqc.
---
do_qc.sh | 35 +++++++++++++++++++++++++++++++++++
parse_fastqc.py | 22 ++++++++++++++++++++++
2 files changed, 57 insertions(+)
create mode 100755 do_qc.sh
create mode 100755 parse_fastqc.py
diff --git a/do_qc.sh b/do_qc.sh
new file mode 100755
index 0000000..54956f9
--- /dev/null
+++ b/do_qc.sh
@@ -0,0 +1,35 @@
+#!/bin/bash
+
+# http://linuxcommand.org/wss0150.php
+PROGNAME=$(basename $0)
+
+function error_exit
+{
+# ----------------------------------------------------------------
+# Function for exit due to fatal program error
+# Accepts 1 argument:
+# string containing descriptive error message
+# ----------------------------------------------------------------
+ echo "${PROGNAME}: ${1:-"Unknown Error"}" 1>&2
+ exit 1
+}
+
+lib=${1}
+report_dir="${lib}_fastqc"
+
+if [ ! -e ${report_dir}.zip -a ! -e ${report_dir} ]
+then
+ cmd="fastqc ${lib}.fastq.gz"
+ nice -n 19 ionice -c2 -n7 ${cmd} || error_exit "${cmd} failed"
+fi
+
+if [ ! -e ${report_dir} ]
+then
+ unzip ${report_dir}.zip
+fi
+
+parse_fastqc.py ${report_dir}/fastqc_data.txt \
+ > ${report_dir}/${lib}_overrepresented.fasta \
+ || error_exit "fastqc result parsing failed for ${lib}"
+
+exit 0
diff --git a/parse_fastqc.py b/parse_fastqc.py
new file mode 100755
index 0000000..079a158
--- /dev/null
+++ b/parse_fastqc.py
@@ -0,0 +1,22 @@
+#!/usr/bin/env python3
+
+import sys
+
+with open(sys.argv[1], "r") as fqc:
+ line = fqc.readline()
+ assert line.startswith("##FastQC")
+ while not line.startswith(">>Overrepresented sequences"):
+ line = fqc.readline()
+ seq, count, percent, source = fqc.readline().strip().split("\t")
+ assert seq == "#Sequence"
+ #cumul_percent = 0
+ line = fqc.readline()
+ order = 0
+ while not line.startswith(">>END_MODULE"):
+ seq, _, percent, _ = line.strip().split("\t")
+ order += 1
+ #cumul_percent += float(percent)
+ print(">Over_represented_%d (%s)\n%s" % (order, percent, seq))
+ line = fqc.readline()
+
+sys.exit(0)
--
GitLab