Peak separation based on the distribution of errors

If the poly(A) evidence errors are normally distributed, use the variance of this distribution to estimate the probability of pieces of evidence at a distant base belonging to the second peak,