Hello everyone and thank you developers for the amazing tool! I am encountering an issue where I am losing majority of the reads at the barcode correction step. Has anyone encountered the problem before? Sounds to me like a barcode sequence issue. I would greatly appreciate any pointers! 🙂

Hi @Barbara Puzek: Thank you for the kind words 🙂 Regarding your question. The one time we have seen this is when the read qualities were low (although usually the reads are also dropped in earlier steps in a steady fashion). Could you send us one of the NanoPlot QCs here? If the quality is good then we will need to investigate further. But before doing so, I also wanted to confirm that this is indeed 10X Genomics prep?

Yeah just to add to Lara's point, we've seen this issue occur before when the barcode format does not match the expected 10X format resulting in that massive drop

Thanks a lot both for your replies, yes it is 10X genomics prep, and we did similar experiments before with the same chemistry with good results. @Austyn Trull were you using 10X prep in this instance or another kit? The read quality is good, see e.g. this post-extract nanoplot qc plot:

Got it, did you run those other experiments through scnanoseq as well? Just trying to gauge if its specific to this dataset or if its been an overall issue Yes, the data was using the 10X prep. I think an issue occurred either during prep or sequencing, we never heard back what the issue was, but we could visually see that the barcode sequences had been malformed in the FASTQs (lots of repeating adapters, etc.) and that resulted in a massive drop in reads.

thanks, yes we ran those other datasets through scnanoseq and they look good with no drop going from extracted to corrected reads I should also add that we have corresponding short read sequencing data that looks good for all datasets (including this one) I will annotate the adapters in some fastq reads that pass and fail the correction and post them here, thanks!

Great thank you, and what kit were these by chance?

10x 3v3

Would you mind sharing the blaze whitelist?

thanks for all your help and time, turns out this was indeed a 3v4 prep, your intuition was correct 🙂🥳

Oh great, glad you were able to figure it out! Let us know if you do end up having any issues down the line!

Great - glad it all worked out. Good luck with the rest of the analysis! 🎉