How does vaf cutoff influence the result?

[EvaPang2022]

Hi,

I ran vardict for three times setting individual allele frequency -f as 0, 0.01 and 0.1.

In the output vcf file, the number of records for these three runs are as follow:
-f 0: 492726
-f 0.01: 455223
-f 0.1: 187461

At the begining, I thought that only those mutations with vcf>cutoff were shown in the vcf file. But I found that in the three vcf files, there were still mutations with vaf less than the cutoff and they were marked as 'f0.1/f0.01/f0' in the FILTER column.

May I know what accounts for the different number of records?

Thanks,
Eva