I would like to perform a pangenome analysis of the Bacillaceae family. I chose 3 close to each other genus (Bacillus, Geobacillus, and Anoxybacillus). I used Proteinortho with default thresholds and rising parameters of coverage and identity (50, 65, 70, 75). What I cannot understand, why in the case of low identity and E-value I didn't get core orthologs, but with high identity and coverage, I obtained about 140 core orthologs for 105 genomes. In accordance with the phylogeny, they are close genus. Could somebody explain to me, why with strict parameters the ratio of the core is increasing?