Yes, you can. When using music calc-covg, you can use the tumor BAM as both tumor/normal, since it's only used to measure sequencing coverage. For the MAF, you need to be careful about recurrent false-positives, especially germline mutations. There are many strategies to filter out sufficiently common germline calls, if you don't have matched normal tissue per tumor. For example, you can use the maf2maf script to generate a new MAF with VEP annotations, and then filter out sufficiently common SNPs/Indels at >0.05% minor allele freq in any EVS/1000G cohort:

perl -a -F'\t' -ne '$cm=0; map{($al,$fr)=split(/:/); $cm++ if($fr>0.0005)}@F[76..82]; print if($cm<1)' input.vep.maf > input.vep.filtered.maf