Normalisation approaches for targeted panels

[ARChakravarthy]

Hi there!

I've been interested in estimating the contribution of mutational signatures defined originally with whole-genome data to variant calls from panels such as MSK-IMPACT ; I wonder what the best normalisation would be (I understand the limitation that low mutation counts may impose on the robustness of the estimates) - would the ratio of occurrence on the panel to the ratio of occurrence in the genome for each trinucleotide be most ideal?

Has anyone in your knowledge benchmarked estimates from mutations reduced to regions that are included in panels versus those called from whole exomes? I'm happy to give this a go if I am not reinventing the wheel.

[shenhaizhongdechanrao]

Dear ARChakravarthy:
Hi, I am also trying to get signatures from some regions of special genes. So do you have to find some ways to solve it?