How to demultiplex Miseq undetermined files?

Question

[siyofuerapajaro]

How to demultiplex Miseq undetermined files?

Hi all,

I had some problem with my data, the results of Miseq illumina are in two big files demultiplex labeled as undetermined files.

I used the barcode in header, data after Illumina processing contains still indexs in header or were eliminated?I´m not sure if i can rescue my data using barcode...

Thanks

[Ahill]

The Undetermined .fastq files contain reads where the demultiplexer can't match a barcode in your Samplesheet.csv - so those reads can't be demultiplexed. Might depend on your workflow or application, but the indexes found by the demultiplexer are typically reported in the .fastq headers, and can be extracted and counted using a shell script (for example). Comparing the indexes reported for the 'Undetermined' reads in the headers versus the indexes in your sample sheet can tell you if there is a mismatch between the indexes in the sample sheet and your library (in which case you would correct your SampleSheet.csv), or some other reason why demultiplexing failed. See previous thread here.

[siyofuerapajaro]

Thanks Ahill!!

The point was that I send a wrong barcode to the lab, so for that I do not know what samples belong to which individual. So, I will try with a shell script or better FastQC?

Thanks again!

[GenoMax]

If you can't reliably say which individual sample should have which index then you are out of luck. If you know the correct sample_ID <-> index association it can be easily corrected in the samplesheet and the core can re-run your demultiplexing for you. This would be the preferred way rather than you trying to do anything.

[siyofuerapajaro]

Thanks genomax! I have the right "sample_ID - index" association. What do you mean with "the core can re-run ",the core Illumina's software? I´m just looking for the better way to rescue them using the "good one barcode". I will try to resolve it using deML.

[GenoMax]

Correct. It would be easiest to have the core re-run illumina software.

I don't understand what you mean by good one barcode? Do you mean that rest of your data is already demultiplexed and there is one sample in "undetermined" because you failed to provide the index for it at submission time? If this is the case you can also use demuxbyname.sh from BBMap suite.

Note: Please use ADD COMMENT/ADD REPLY when responding to existing posts to keep threads logically organized.

[siyofuerapajaro]

About good barcode. Sorry, I was not clear enough,with good barcode I mean the correct indexes-samples list (I failed to provide the correct indexes for it at submission time).

Most samples are in the undetermined files.

Thanks again genomax for your help, I will try them.