All reads deposited in Undetermined indices

Question

[EVR]

Hi All,

I have a RNAseq sample for which after demultiplexing I found almost all the reads got deposited in 'Undetermined indices'. So can I take these fastq files for further comparison or do I have to check for quality of the sample.

Regards

Dino

Answer 1

[Devon Ryan]

You should look at why this happened. If you just had a bubble go through during the index read then just increase the mismatch allowance in the barcode and rerun things. If you misspecified a barcode then fix the sample sheet and rerun. This all assumes that you actually did have multiplexed samples, of course (if not, then go ahead and use the undetermined indices files).

For reference, using only perfect matches I typically see only 1-2% of the reads end up in the undetermined indices fastq files (allowing a single mismatch drops this to <1%).