Use mutect2 or UnifiedGenotyper?

Hi,

I am currently working on finding the somatic mutations in the tumor (using the sing cell sequencing result). And my final goal is to use these somatic mutations from different regions to build phylogenetic tree for one tumor.

In most papers, they use UnifiedGenotyper to call mutations. Then they acquired the somatic mutations by removing the SNV sites called from the normal sample. As the result is a joint calling result, each sample has all information on all sites, which means it can be easily used to build the phylogenetic tree.

As Mutech2 is designed to call somatic mutations, it may have better ability to do somatic calling. If I merge the Mutech2 result together and check whether the empty value is due to having no variant or having a gap, I can also acquire the vcf results, in which each sample have all information in all sites.

My question is that can Mutech2 result be used in this way? What is the difference between the somatic mutations selected from the snvs called by UnifiedGenotyper and the somatic mutations called by Mutech2?

Thanks for any help.