How to create a list of the barcodes located in the read headers of demultiplexed fastq files?

Question

[DVR]

Hello all, is there an already available script to obtain the list of the barcodes embed in the header of the reads from my multiplexed fastq files? Thanks a lot!

[h.mon]

Can you show the first 4 lines of your fastq? Either

head -n 4 file.fastq

Or

zcat file.fastq.gz | head -n 4

The solution will be something like

zcat file.fastq.gz | cut -f 2 | sort | uniq

But to know exactly, we need an example from the header.

[DVR]

Sure! here they are! Thank you!

@M01380:50:000000000-AV1DH:1:1101:13660:1636 1:N:0:M154:16S_V1V3 TTCTGCCT|0|TGAACCTT|0 CS1_534R_YM3_for|4|27|
GATGAACGCTGGCGGCGTGCTTAACACATGCAAGTCGAACGGGATCTGACCAGCTTGCTGGTTGGTGAGAGTGGCGAACGGGTGAGTAATGCGTGACCAACCTGCCCCATGCTCCAGAATAGCTCTTGGAAACGGGTGGTAATGCTGGATGCTCCAACTTGACGCATGTCTTGTTGGGAAGGGGTTTTGGGCATGGGGTGGGGGTGGGTCCCTTCCGGCTTTAGGGGGGGGTATGGGCCACCTTGGCCTTGGTGGGGGACCCGCCTGAGGGGGG
+
GFFGGG<FFEGGGGGGGGGGGGFGGGGGGGGGGGFGGGGGGGGGGGGGGGGEGGGGGGGGCECEGGGGGFFGGGGGG@=FGGCF+BFFGGGGEGGEGDDFFG@DDGGECGFG9FGFFGGGFGGGGGGFGGGFGGGGFEG<,DFFAFFGFGGGFCG:<DDGG9FG*:CE9CEF?EFF>*:>****3C7***3***3A8*2/:8C**1*)/21***0.9))7/)*+2*)7>C))1)0C)*97)74)*0)0*).)*(07)(((((0(((((((-(75

[DVR]

Just a clarification: I don't want to cut them out from the header but just have a list of them by file (sample). Thanks for your good disposition!!

[Joe]

Please dont post answers instead of comments, unless answering the question.

Also to ensure readability of posts, please use the code markup button around and code out input/output (the button with 101010 on)

[h.mon]

zcat file.fastq.gz | cut -f 3 | sort | uniq -c

It is essentially the same answer as doctor.dee005, but mine counts the number of occurences of each barcode (uniq -c), and his is more elegant as it considers only header lines.

I am assuming the barcodes are the TTCTGCCT|0|TGAACCTT|0 part.

**edit: probably the cut command has to be:

cut -d " " -f 3

Answer 1

[Dattatray Mongad]

I suppose your header are like this:

@7001367R:585:HNHVHBCXY:1:1102:1267:2073 1:N:0:ATTACTCG+TATAGCCT

where HNHVHBCXY is your barcodes. If your fastq files are in gzip format. Do following:

 zcat file.fastq.gz | awk 'NR%4==1' | cut -d ':' -f3 | uniq

[DVR]

Just a clarification: I don't want to cut them out from the header but just have a list of them by file (sample). Thanks for your good disposition!!

[Joe]

Pretty sure that is not the barcode (though I might be mistaken), according to Illumina, that's the flowcell ID (though in principle the approach will work if you change the column number):

http://support.illumina.com/content/dam/illumina-support/help/BaseSpaceHelp_v2/Content/Vault/Informatics/Sequencing_Analysis/BS/swSEQ_mBS_FASTQFiles.htm

Answer 2

[drkennetz]

When you say a list my mind is immediately going to python but you haven't tagged it so I will give an awk solution for you:

zcat file.fastq.gz | awk 'NR == 1 || NR % 4 == 0 ' | awk -F ':' '{print $12}

Your fastq is a little weird in the header line, but This will print the first line, and every fourth line (which should be header if it is like header, sequence, +, quality ( the NR % 4 == 0) bit may need to be changed to NR % 5. Then it will set the field separator to : and select the column with the index info, but you will also pick up some stuff on the back end too. You will definitely get indexes!