Chimera removal --p-min-fold-parent-over-abundance NUMBER query

User Support
1

Hi, I had a query about altering --p-min-fold-parent-over-abundance during denoise-paired using dada2.

The default is one and with this I lose just under half my reads. I realise I can increase this (maybe to 2?) but this will increase the likelihood of chimeras in my dataset and I don't want to do that. My PCR has 35 cycles as I have a low abundant target, so I expect more chimeras, so maybe 1 is just doing it's job fine! It seems more intuitive that there should be more parent than chimera so maybe a bit above 1 would be realistic. I don't really know how to test this as there will be know knowing if they are chimeras or not. I'm just after some insight or reassurance really!

Thanks!

2

Hi @SarahH!

Where are those reads being lost? Can you please share your denoising stats?

3

Hi!

I just had a query regarding the interpretation of stats-dada2.qzv with regard to the denoised column during denoise-paired in dada2.

I only have a very small proportion of reads lost here, but I just wanted to understand why they had gone, as they had obviously passed the filter step. Is it that an error had been detected but not been able to be resolved so it drops the read?

Many thanks

4

Hi @SarahH, I merged your two posts together, since they seem to be going about asking the same question from different angles.