Suggestions for dada2 trunc parameters

#1

Hi,

Can anyone offer a suggestion for what my trim-left and trunc-f/r parameters should be based on this quality plot?

I’ve attempted to do @120 for both fwd/rev and @150 for both fwd/rev. Didn’t trim-left at all yet. I am running out of ideas and would like a new perspective for my next run.

If anyone has any suggestions and a tip for reading these plots to determine DADA2 parameters, it’d be much appreciated.

I am confused about how significant quality score drops are. Scores do drop from 37 consistently to whiskers at around 25 after a while but is that significant enough to truncate from there? Should I immediately truncate as soon as there is a drop, regardless of how significant/non-significant the drop is? That is the trend I’ve been observing in terms of suggestions – if the score drops, truncate.

Thank you so much!

0 Likes

(Matthew Ryan Dillon) #2

This is very strange looking data. Can you tell us more about your sequencing platform, the target amplicon and region(s), and any preliminary QC that has been performed on these reads? Thanks! :qiime2:

0 Likes

#3

@thermokarst

Thanks for the response!

Sequencing platform: iSeq 100
Target: 16s rRNA
Regions: V3, V4

Not sure about the preliminary QC that was performed on these reads.

What are you observing from the unusual plots?

Thanks for your time!

0 Likes

(Nicholas Bokulich) #4

The quality scores are all very high and quite uniform. This does not look typical (e.g., search this forum for other topics asking about dada2 trunc parameters to see what typical quality profiles look like). This probably relates to the sequencing platform — iseq100 is quite new and we have not seen data from this platform reported on this forum yet.

That’s another strong possibility that would explain this profile.

Based on your results, though, I’d recommend proceeding without filtering because the Q scores are all very high.

0 Likes