comparing the single-end and paired-end output at demux step

I have got another trouble in demux step. The sequences count of paired-end does not match with paired-end one. Could you please share your idea?

paired-end result:

single-end result:

*paired-end output looks weird, I do not know why. In contrast, single-end is much better than paired-end. What probably makes this issue?

Thanks friends

Hello there @TurboQiimer!

Neither do we! You haven’t provided any info about what you have done, or how you have done it. Can you please provide more details, including information about the data you are attempting to work with, how you imported it, and any other relevant details? Thanks so much! :t_rex:

Thanks to you sir @thermokarst

I am working on treated samples in two time points plus their controls which including three replicates each, altogether 12 samples. The sequencing platform is Ilumina HiSeq. I imported the sequences in two ways: 1-only forward 2-both forward and reverse. As you see the photos I shared here.
I used barcodes it means my samples are multiplexing that’s why I applied –type Multiplexed, so I needed to demultiplex them at first. The commands from importing to demuxing are as follows:

importing (single):
qiime tools import
–type MultiplexedSingleEndBarcodeInSequence
–input-path forward
–input-format MultiplexedSingleEndBarcodeInSequenceDirFmt
–output-path forward.qza

Importing (paired):
qiime tools import
–type MultiplexedPairedEndBarcodeInSequence
–input-path Both
–input-format MultiplexedPairedEndBarcodeInSequenceDirFmt
–output-path Both.qza

demuxing (single):
qiime cutadapt demux-single
–i-seqs forward.qza
–m-barcodes-file metadata.tsv
–m-barcodes-column staggerandbarcodeF
–o-per-sample-sequences forward.qza
–o-untrimmed-sequences forward.qza

demuxing (paired):
qiime cutadapt demux-paired
–i-seqs both.qza
–m-forward-barcodes-file metadata.tsv
–m-forward-barcodes-column staggerandbarcodeF
–m-reverse-barcodes-file metadata.tsv
–m-reverse-barcodes-column staggerandbarcodeR
–o-per-sample-sequences demuxb.qza
–o-untrimmed-sequences demuxboth.qza

The I visualized them by this command, and shared their results:

qiime demux summarize
–i-data demuxb.qza
–o-visualization demuxb.qzv

Now, I have no any clue why the sequence counts are hugely different in two methods. I stopped analysing at this earlier step to fix the issue then move into the next step. It is a baffling case to me indeed. I was on vacation maybe I forgot something :grin:

If you need to ask questions, please do that. I am vigorously eager to solve this issue cus that is quite crucial for the downstream steps and also has a direct influence on my result. Finally, I want to know why the single-end and paired-end demux results are not similar, and what makes them different? Or I need to know what is my mistake?

Thanks a bunch

1 Like

Dear @thermokarst and other supporters,
I am still blocked at this step over one week. Please give me a hand.

Judging by your demux-paired command, it looks to me like you have dual-index barcodes. If that is the case, it is impossible to demultiplex just the forward reads - there simply isn’t enough information available to successfully demux. Please confirm your barcoding strategy. Thanks!