DADA2 p-n-reads learn

[Jibda (Jibda del Pilar Carreño)]

Hi everybody

We are working with Dada2 but I have two questions related with one of the parameters in the process of Denoise and dereplicate paired-end sequences --p-n-reads-learn about the The number of reads to use when training the error model, :

1. The number of reads that the program takes for make the error model is per sample or are the total reads I have entered.
2. If I don´t have 10000000 of reads by default, the program would take the greatest number of reads to make the error model with out problem or I have to change that parameter with the greatest number of my reads?
   Thanks!!!

[ebolyen (Evan Bolyen)]

Hi @Jibda!

Sorry for the very late response.

@Jibda

The number of reads that the program takes for make the error model is per sample or are the total reads I have entered.

Total reads encountered, I believe the order of those reads is arbitrary, but the error profiles should be consistent within a run.

@Jibda

If I don´t have 10000000 of reads by default, the program would take the greatest number of reads to make the error model with out problem or I have to change that parameter with the greatest number of my reads?

It will use all of your reads if you have less than the default (or whatever value you might pass).

Hope that helps!