Shared peptide quantification

[rsalz]

Hi, I have a more conceptual question about the shared peptide quantification. I am using metamorpheus and then running flashlfq after (since bayesian fc analysis isn't part of metamorpheus quant yet). I was wondering if by turning off shared peptide quantification, abundances of shared peptides are not used in protein inference/quant? Are those thrown out completely or does it default to 1 of the proteins that it's shared among? Is there no concept of protein "groups" as a happy medium between those 2 options? I have a lot of proteoforms in my database and I get strange results when I use the shared peptide quantification setting but I don't want to throw valuable information away.