I am looking at my quality data after importing casava 1.8 R1 and R2 files. The forward reads maintain their high quality for a couple hundred basepairs, whereas the reverse reads maintain the quality for a few dozen base pairs less. Should I truncate the forward one sooner just because the reverse one is low quality?
For example, at base 150, the forward is high quality and reverse is low quality. Should I truncate both at 150, or just the reverse one?
Thanks for your time.
Hi @msport469,
You can absolutely truncate the forward and reverse in different places if there's a different drop off in quality. I do it often when my forward reads are high quality and the reverse reads have a lower drop in quality. In fact, this is a common behavior in Illumina data.
There is one caveat: If you trim your reverse data too much, you may no longer be able to overlap enough to join the ends. So, just watch your trim length if you want to join reads.
Best,
Justine
P.S. I moved this to user support, since it seems like a more general question. (Community plugin support is where you can go ask about non-core QIIME plugins like q2-PICRUSt).
Thanks for getting back to me so quickly,
What are the commands to do so? What I tried (below) didn't seem to work.
qiime dada2 denoise-single
--i-demultiplexed-seqs '/scratch/msportie/uganda/scripts/demux-16s.qza'
--p-trim-left-f 0
--p-trunc-len-f 250
--p-trim-left-r 0
--p-trunc-len-r 144
--o-representative-sequences rep-seqs-dada2-16s.qza
--o-table table-dada2-16s.qza
--o-denoising-stats stats-dada2-16s.qza
The error I got was associated with the -r and -f parts not being recognized (I no longer have the error message it gave me. my apologies).
Thanks for your time.
Hi @msport469 --- without the error it is a little hard to say what went wrong, but my guess is that you are trying to use parameters that don't exist on the denoise-single command. Specifically, single-end reads don't make a distinction between forward and reverse orientation, so there aren't parameters there to control for trimming forward and reverse reads separately. Make sense? You mentioned about that you had R1 and R2 files, which implies you have paired-end reads, so there is a good chance you just need to use the paired-end equivalent command: denoise-paired.
Wow thank you. On a very similar not, I have another set of files that is also paired that I accidentally did denoise-single on, but the noise level is identical for both pretty much so I don't want to truncate them at different places. Should I stop that job that's still working and resubmit with denoise-paired as well or will it be fine with denoise-single since there is no difference in trimming or truncing?
If you ran denoise-single on paired-end reads, you have told QIIME 2 to throw away all of your reverse reads to make that happen. This means that DADA2 will not join reads (because the reads aren't paired), which means you will be denoising a region that is potentially (much) smaller than your target region.
Okay. thank you for your time.
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.