Does Illumina Undetermined Fastq files contain barcodes

Question

[Paul]

Dear all,

I have some questions regarding demultiplexing Illumina BCL files. First question is if Illumina´s Undetermined Fastq files contains barcodes?

Second question - If I would like to provide demultiplexing and keep my barcodes in fastq files (no in the header like CASAVA provide, but inside the read sequence) - does anybody have any experience how to do that?

Thak you for any sharing your experience.

Answer 1

[Devon Ryan]

Yes, the undetermined files have the barcodes in the header of each read.

I haven't a clue how you would include read #2 into read #1, though perhaps playing with --use-bases-mask would do the trick (presumably by adding the index length to that of read1). It would make more sense to me to just write read 2 as separate file (so paired end experiments will have all three reads rather than just two).

[Paul]

Thank you Devon for answer.. I have Undetermined Fastq files from MiSeq (MSR does not print barcode in the header like for example bcl2fastq).

this is my first read of undetermined FASTQ file:

@M03456:4:000000000-AGW03:1:1101:15913:1332 1:N:0:0
ATACAGACATATCTGTACGTGAACAATGCTGGTCCCTGTGGTTTGCCCACCTGCATCTATGCATTTGTTCACGTGGATGTACCCAAGTCCCTGGTCAGACCAGGAGCACCTCAGTGGCTCCTGGGAAAGAGGAAAGATCGGAAGAGCACAC
+
3>AABFFF4@D@GGFBGG4BF4FGBFBAEGHBC5AFFGFFFGCG?HH3FGECHAGHBGHHDGFDFFFFFGHGGGHAGHGHHEHFDEEGFHHHGHGGFGHHHGHEEEEAHGHHHGGFHGHFAGFCE0CGFFGE0B3B33FFADF?0BG1CGH

I am not sure how MiSeq Reporter works when creates Undetermined fastq files.

Thank you.

[Devon Ryan]

Ah, I never have to handle our MiSeq data, sorry.

[Ido Tamir]

why don't you use bcl2fastq?

[juan_lu_97]

were you able to fix it? I have the same problem...

Answer 2

[ptinto]

MiSeq MSR does not put indexes in header but the index_rank (row). If the index used is not found in the sample sheet, it put a 0 (the last :0 in the header of the undetermined_reads.

With bcl2fastq you can ask to have the indexes written as a FASTQ with option `--create-fastq-for-index-reads

You can set MiseqReported to produce the fastq of the indexes modifying the MiSeqReporter.config with:

 <appSettings>
            <add key="CreateFastqForIndexReads" value="1" />
 </appSettings>

If you want to explore the indexes in the undetermined to see if you have mispelled an index in the sample sheet and you are loosing this sample in the demultiplexing, there is a file in the flowcell folder where are recorded all the index found an how many tiems they where found (you can acces this data also from the MSR) but I can not remember the path right now (this file is different in MiSeq and in NextSeq500).

[Sus]

The path for the file is <run_directory>/Data/Intensities/BaseCalls/Alignment/ and is called DemultiplexSummaryF1L1.txt

(I know the post is 18 month old but it could help someone in the future)

Answer 3

[Ido Tamir]

1. yes (but simply try it out)

2. FASTQ is a shitty format and there is not way to identifiably represent the index read in the read sequence (beside the name field). One possibility would be to use a third or 4. file: --create-fastq-for-index-reads. Another would be to switch to BAM files, where you can add the index reads in tags.