discrepancy between alt_frac and coverage methods

[igordot]

I tried running alt_frac and coverage scoring methods on the same inputs. Then I calculated the VAF based on alt and ref reads from the coverage output matrices. Most of the values agree, but some of them have minor discrepancies. I assumed they would all be the same. Is there some additional filtering that either of the methods performs?

[gloriafight]

Can you show your code about calculate coverage? I want to do this,but the results were strange. My code is "vartrix_linux -v $vcf -b $bam -f $fa -c $tsv -s coverage --threads 45 -o out/3d.coverage.mtx". And I want to achieve this goal "To build phylogenetic trees based on the mitochondrial mutations, we retained mitochondrial genome position with coverage > 20 and at least 85% cells within the interest cell populations expressed. Because of heteroplasmy of mitochondria genome, a cutoff of alternative allele frequency > 0.1 was set to identify subclone mutations in the mitochondrial genome."