I run the analyses on two distinct batches of samples and then I merged the OTUs tables.
I found out a huge loss of information in terms of the number of OTUs and the number of sequences per sample compared to the OTU table from each distinct run.
I tried then to merge the sequences (cat seqs.fna, seqs.fna> merged.fna) and the mapping files after demultiplexing and run the whole downstream analyses.
I gained the same results of merging the OTU tables at the end of the two individual runs.
Does anybody know what is due to this loss of information and how to avoid this?
thanks, everybody for your kind attention and help!