How to make assnake to use the original raw data, not preprocessed at previous time?

[PollyTikhonova]

Situation: I run assnake result dada2-filter-and-trim --preset def -d $DNAME gather -j 50 --run
It returned the result I was not pleased with. I changed a preset and run again. Then, I realised, that assnake used not the original raw data, but a preprocessed one by the first run.
To my mind, it is not very obvious behaviour. But I am interested is there an option to use the original raw dataset (or maybe an option to outline a name of preprocessed version of dataset (a folder name), for example :
raw, or raw__dada2fat_def.fbe5687 or raw__dada2fat_def.fbe5687__dada2fat_manual

[PollyTikhonova]

Actually, I found some logic inconsistency:
before I run assnake result dada2-filter-and-trim, I run assnake result dada2-full. It failed, then I used dada2-filter-and-trim. After that, I run again assnake result dada2-full considering that like at the situation descibed in my previous comment this script will use the newest (filtered and trimmed) dataset. But it wasn't.