hi everyone
my data paired-end Miseq reads and I use from 27f/534r of V1-V3 from 16 s.quality of my data is not very good. what is your suggestion for denoise? I don’t know what to write for p-trunc-len or --p-trim-left in dada2 method. first, I search in the forum but don’t use anyone from my primers. please help me.
which primers you use does not matter — the same advice applies from other forum posts for trimming your data. As a rule of thumb, I recommend truncating your sequences where the mean Q score drops below 20. If you share the output of qiime demux summarize we can offer more specific guidance.
please give me answer to my question. i have a lot of problem in denoising. this section has very important for me because my read are low quality.
Hi @elaheahmadi,
Please read the forum code of conduct, particularly this section. We are very happy to help, but are volunteering our time. We field many questions at the expense of our own time.
See my answer above;
So I would recommend:
- Truncate the forward reads to ~140
- The reverse reads are really low quality; I would recommend discarding these and only using the forward reads. If you must keep them, truncate the reverse reads to ~120 if that gives you enough overlap between forward/reverse (you need a minimum of 20 nt to overlap successfully).
You can experiment truncating the forward/reverse reads at different sites to see how this impacts your sequence yields — truncating less so that you have longer reads will result in more error and a greater chance the read will be dropped. So you can push the envelope a little bit to see if you can extract enough longer reads to proceed.
Good luck!
1 Like
Dear Nichlas
thank you so so much your answer.
Best
This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.