Reads which do not map

I followed ‘moving pictures’ tutorial with my 16S data.
The result of DADA2 was a Feature-Table and Feature-Data.
I expected that the total number of input reads would be the same with total number of frequencies. but it wasn’t.
Is there any way that I can get reads which do not map in each sample?

Thanks
Choe

Hi @Choe,
You shouldn’t expect the total number of reads to be the same before and after DADA2. Dada2 is going to be discarding several categories of reads during its denoising process. For example: reads considered to be chimeras, reads with too low quality scores or too short, reads that identify as PhiX, and singletons, to name a few. The number of reads that are discarded largely depends on your data so there is no real way of estimating how many of your reads should be retained after DADA2. That being said you can share with us the ‘stats summary’ and we can perhaps evaluate if the results are consistent with what we expect or not.

Not in qiim2, perhaps in the R stand alone version of DADA2, however, generally these reads aren’t biologically relevant so I’m not sure if these ‘discarded reads’ would be of any use for you aside from a sanity check.
Hope that helps!

1 Like

Thanks!
It helps a lot :slight_smile:

1 Like

This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.