SplitBarcode

The repository provides scripts for spliting PE fastq from MGI sequencer platform by barcodes sequence.

Author

The current implementation was written by caoshuhuan (caoshuhuan@yeah.net). I would appeciate if you send email to me when you have any question about this script or report bug !

Version history

The current code version is v1.0.2

v1.0

• split PE fastq with single barcode
• outputs are compressed
• some statistical results provided

v1.0.1

• delete parameter -l
• modified compress method to reduce process time
• submit SplitDualBarcodes.pl for MGI dual barcodes multiplexing, type in perl SplitDualBarcodes.pl -h to get the tutorial of this script

v1.0.2

• support SE data

v1.1 (under deverlopment) :

• intergrate single and dual barcodes multiplexing module
• support windows system

Prerequisites, Tutorial and Results

System: Linux
Memmory: 1 Gb or above
Storage: 1 Tb or above
Perl version: 5.16 or above

The script runs on CentOS 7 or other linux systems on a 64Bit machine with Perl 5.26, for 100Gb data, It will take about 2 hours with 1 Gb memory if final fastqs uncompress.

Tutorial

Usage:
	perl splitBarcode_PE.pl [options]
		*-r1 --read1 <string>		read1.fq.gz
		 -r2 --read2 <string>		read2.fq.gz if not provided, it will be SE
		 -e  --errNum <int>		mismatch number [default: 2 for PE, 1 for SE]
		*-f  --firstCycle <int>		First cylce of barcode
		*-b  --barcodeList <string>	barcodes list
		 -rc --revcom	<Y|N>		generate reverse complement of barcode.list or not [default: Y]
		 -c  --compress <Y|N>		compress(.gz) output or not [default: Y]
		 -o  --outdir <string>		output directory [default: ./]
		 -h  --help			print help information and exit

• * means parameter must be provided.
• the default mismatch value is 2.
• the default output directory is ./.
• the fastq will be compressed in .gz format when -c Y has been set and run gzip.main.sh after split process finished.

Command line example

perl SplitBarcode.pl -r1 read1.fq.gz -r2 read2.fq.gz -e 1 -f 101 -b barcode.list -r N -o /path/outdir -c Y

Please make sure the first cycle number of barcode correctly.

barcode list example

Only barcode name and barcode sequence need, seperated by tab or space.
barcode will miss if lane starts with #, for example: #96 ATGATCTAGC.

1	ATGCATCTAA
2	AGCTCTGGAC

The default barcode sequence is a reverse complement of sequence between first cycle and last cycle in Read_2 fastq file. If not, set parameter -rc N.
for example:
if one read from read2 fastq is:

@V300000000L1C001R001000000/2
TGACTCAATCATACGTTTATACCTCCTATAGTAAAAAGTTTTGTCTTCTTTCAGATATAAGTGTCTCTGTGATGCAGGCTGGGTTGGCATCAACTGTGAATCATTCCAAC
+
FEFGEGGGFGGEEGEFGEEEEBGEFDEEGDBGEGEEAFFGGGDGFEEEEFEFFGFGEFCGDEEFGGEFEEECGBEDEGFFDFFEFEGDGGFFE?EEDCFF71,'962'&)

the barcode sequence in read 2 is TCATTCCAAC,
the read can be splited perfectly when the barcode provided is GTTGGAATGA or TCATTCCAAC

Output

There are several types of file generated after script finished：

• barcode_1.fq(.gz), barcode_2.fq(.gz)
• BarcodeStat.txt
• TagStat.txt

- barcode fastq

The format of fastq name is:

Chipname_lane_barcode_1.fq.gz : V300000000_L01_1.fq.gz
Chipname_lane_barcode_2.fq.gz : V300000000_L01_2.fq.gz

Chip name and lane name are captured from the read1.fq.gz. Also there is a couple of fastq named undecoded_1.fq.gz and undecoded_2.fq.gz, to keep reads which don't contain any barcode sequence.

- BarcodeStat.txt

BarcodeStat.txt counts the reads number and barcode split ratio of different barcode separately. In finally, the Total lane calculate the total reads number and ratio.
The format of BarcodeStat.txt

#SpeciesNO	Correct	Corrected	Total	Pct
1	95327109	4112238	99439347	19.2152%
2	93797238	6267736	100064974	19.3361%
...
Total	468560368	27305422	495865790	95.8187%

column	name	description
1	SpeciesNO	barcode name
2	Correct	the number of reads without mismatch
3	Corrected	the number of reads within mismatch value
4	Total	Correct and Corrected reads number
5	Pct	percentage

- TagStat.txt

Tag means a short sequence locate between the first cylce and last cycle on Read 2 fastq. TagStat.txt exhibit all tag number and percentage.
The format of TagStat.txt is:

#Tag	SpeciesNO	readCount	Pct
ATGCATCTAA	1	99439347	19.2152%
...
ATGATCTAGC	unknow	200	0.0000%

column	name	description
1	#Tag	tag sequence
2	SpeciesNO	barcode name or unknow
3	readCount	reads number
4	Pct	percentage