Why there is a 50% variant allele frequency in TCGA NSCLC data?

Question

[sm.hashemin]

Hi, I was taking a lot at the variant allele frequencies in TCGA NSCLC data (from the the MAF somatic SNV file). I have seen thet the highest VAF is 50% and it apears that they have a cutoff there. Does it mean that a SNV of 60% vaf can not be a homozygous mutations in 60% of cells (while tumor purity of 60%) or a subclonal snv at higher purities?

As well there is a lot of variants with 0% VAFs! which is also strange.

What I want to learn from this story is: what should I consider when facing a oncoplot? can a variant on oncoplot have only 3% vaf? specially in paired primary vs recurrent tumors, can it be that a variant is simply missing in one because it has a less than cutoff level vaf ?

Best

Answer 1

[sm.hashemin]

Answering myself: because mutations are considered to happen in heterozygous state, therefore at 100% purity you will have a 50% VAF. VAF of more than 50% belongs to the LOHs.