Multi samples (Normal and Tumor) somatic SNPs identification

Question

[Omics data mining]

Hi everyone

I have bam files of multiples normal and tumor samples. For the efficient SNPs prediction, Its recommended to have merged multi sample vcf file. These are steps I followed

1. Mapping of all T and N samples to get multi samples bam file ( Tumor 20 Bam files : Normal 20 Bam files)

2. For the VCF detection , there are two options : In my case, I have two groups of patients.

option 1: Run Mutect2 on all the Tumor and Normal samples get multi sample vcf file as given here https://gatkforums.broadinstitute.org/gatk/discussion/12615/using-mutect2-to-call-the-same-variants-in-multiple-tumor-samples

For the SNPs detection, Can I proceed with this multi sample vcf (consist information of both groups) and perform the downstream analysis ?

option 2 : Or I should have multisample vcf of one group of Pateints (group1) , multisample vcf of another group of patients and then do the downstream analysis?

I want to be sure for my analytical approach. I will appreciate all the suggestions.

Thanks in advance