I am running some analyses on low biomass samples.
In may hands up to know I have found out that the barplot calssification is better when data are processed with vsearc quality filter dereplication, ... done after paired end reads joining.
anyhow I would like to optimise also the dada2 pipeline due to suggestion by the forum leaders!
Could I try and run vsearc filetring and dereplication after performing dada2 denoising?
I use dada2 directly on the R1 and R2 raw files
Is it a good idea I should pursue or bad, uncorret?
Another important parameter changing the classification is in
qiime feature-classifier classify-sklearn
--p-read-orientation 'same'
parameter which really changed the situation seen in barplots (I refer to the vsearc filter/dereplication approach on joined sequences.
Hi @MichelaRiba,
Can you share the exact commands you are using in your current work flow? and if you are you seeing any errors please share those as well, please. What are your goals in changing your workflow? What are you hoping to change about your outputs? More details about your goals would be so helpful!
--Hannah
I Will Fw the commands, if really Need, anyhow my question Is theoreticcal, conceptual. As I mentioned I had bad classification with dada2 compared tò vsearc workflow for this reason I would like tò see of something can be done towards using dada2 workflow as most endorsed.
Hi @MichelaRiba,
Yes, please pass along your commands. It will help a lot in understanding your current pipeline so that we can help you better.
--Hannah