bcl2fastq dumps most of the reads as undetermined

Question

[rohitsatyam102]

Hi Everyone. I have Iseq run BCL files and I am trying to perform demultiplexing using the following command

bcl2fastq --runfolder-dir 20221017_FS10000388_34_BPN80013-1437 -p 12 --output-dir 20221017_FS10000388_34_BPN80013-1437/fastq_files --sample-sheet 20221017_FS10000388_34_BPN80013-1437/SampleSheet.csv

I am trying to figure out what went wrong. The multiqc report after running the bcl2fastq can be found here

My Samplesheet looks like:

Answer 1

[GenoMax]

Take a look at DemuxSummaryF1L1.txt file in Unaligned/Stats directory of the run to get an idea of what the "Undetermined" pool of index sequences looks like. Be sure to scroll down for a bit in this file to get to a section that looks like:

### Most Popular Unknown Index Sequences
### Columns: Index_Sequence Hit_Count
TCACCAAG+CACGGCGC       360
TCAACAAT+CACGGCGC       320
TCACCGAT+CACGGCGC       240

Generally this error happens because the indexes have either been reverse-complemented or are plain wrong in your SampleSheet file. Once you fix that issue the demux should proceed fine.

[swbarnes2]

Alternately, you can look at the LaneBarcode html output, which is buried in the reports. The path will be something like

Data\Intensities\BaseCalls\Reports\html\FS10000388\all\all\all\laneBarcode.html

But it will only show the 10 most common unknown barcodes per lane.

[rohitsatyam102]

Yes, @swbarnes2. Thanks for this recommendation. I wasn't aware about this.

[rohitsatyam102]

Hi GenoMax . Thanks for saving the day. Yes it was the i5 indexes (index 2) that needed to be reverse complemented in the sample sheet. I achieved that using this online server and reran demultiplexing. It worked.