When I perform the trimming of my reads after the quality control check, if I have for example sequences 200bp long and sequence 300bp long and I decide to cut at 250 I lose those at 200 bp?
In which file produced by the analysis can I find the number of sequences for sample?
That's just how dada2 does its truncation (I assume when you say trim you really mean the trunc-len parameter) — you can search the forum archive for more discussion on this, but in summary (a) trimming to the same length creates uniform ASVs, improves the error model, and (b) you can always disable this in dada2 and use something like trunc-q instead.