Odd display of demux plot

User Support
, ,
3

Thanks for the welcome!

  1. Only a few QC steps were performed by the sequencing company, including throwing reads with too much adapter contamination. I don’t think this is the issue.

  2. I’ve confirmed that our reads were sequenced with a Novaseq and indeed have only 4 quality scores, 2, 11, 25, 37. I’ve come across this issue raised on GitHub. @benjjneb recommends enforcing monotonicity in the fitted error model by changing the error model matrix. I can comprehend how to do this in R, but I am running dada2 in the qiime framework on the command line and don’t know how to intercept this matrix when running the qiime dada2 denoise-paired step. Perhaps the solution is to export the qiime object and and run the dada2 steps independently, then import the data once denoised.

Guidance would be appreciated!

1 Like