To determine P(m,f), we first assume that P(m) and P(f) are statistically independent, and that P(f) is uniformly distributed (i.e. P(f)=1) and P(m) is a 1/3 of the expected mutation frequency for the studied tumor type (representing equal prior for all substitutions). In practice, we use a typical mutation frequency of 3×10−6 which yields θT=6.3.
