FastDemultiplexer

FastDemultiplexer copied to clipboard

This is a read demultiplexer.

Pros:

• Can manage a lot of mismatches

== Motivation ==

The Illumina HiSeq 1000 dumps .cif (intensity) files, .bcl (base calls) files and .clocs (probably a summary of the intensities?)files, among others.

Basically, for each cluster on the flow cell, there will be 4 sequences: R1, R2, R3, R4. R2 and R3 are the indexes, or bar-codes, and R1 and R4 are the sequences.

There also may be only R1, R2, and R3 (RNA-Seq and ChIP-Seq)

CASAVA 1.8.2 only allows 1 mismatch per index (2 for dual-indexes).

Furthermore CASAVA 1.8.2 has been released many years ago.

This demultiplexer allows more data to be retrieved and is maintained too !

This demultiplexer is implemented in Python as a single executable.

This is GPL work.

== Input files ==

The files can be in fastq or in fastq.gz.

== Output files ==

fastq.gz

== Conversion from BCL/CLOCS to FASTQ ==

#!/bin/bash #$ -N convert-HiSeq1000-BCL-2011-12-21.8 #$ -P nne-790-ab #$ -l h_rt=48:00:00 #$ -pe default 8 #$ -cwd

sequenceWorld=/rap/nne-790-ab/Instruments/Illumina_HiSeq_1000_Hellbound run=111207_SNL131_0065_AC0947ACXX NSLOTS=8

source /rap/nne-790-ab/software/CASAVA_v1.8.2/module-load.sh

configureBclToFastq.pl
--input-dir $sequenceWorld/$run/Data/Intensities/BaseCalls
--output-dir $sequenceWorld/$run/FastQ-Sequences/no-demul-Unaligned
--use-bases-mask Y*,Y*,Y*,Y*

cd $sequenceWorld/$run/FastQ-Sequences/no-demul-Unaligned

make -j $NSLOTS

== Demultiplexing FASTQ files ==

By now, you should have one directory per lane.

[@colosse1 FastQ-Sequences]$ ls no-demul-Unaligned/Project_C0947ACXX/ -1 Sample_lane1 Sample_lane2 Sample_lane3 Sample_lane4 Sample_lane5 Sample_lane6 Sample_lane7 Sample_lane8

To demultiplex lane 7, run this:

 FastDemultiplexer.py SampleSheet.csv 7 Project_C0947ACXX/Sample_lane7 Demultiplexed > stat.txt

This will generate

Demultiplexed/Project_A/Sample_X Demultiplexed/Project_A/Sample_Y Demultiplexed/Project_A/Sample_Z ...

and

Demultiplexed/Undetermined_indices/Sample_lane7

== Author ==

Sébastien Boisvert