Fig. 2: Evidence that mutation rate varies between divisions.
From: Rate volatility and asymmetric segregation diversify mutation burden in cells with mutator alleles
a The negative binomial as a gamma-Poisson distribution. The gamma distribution takes the same shape parameter (θ) as the negative binomial and describes the variation in the rate parameter (λ) of a continuous mixture of Poisson distributions. b Schematic of single-cell lineage showing summing of segregant groups to determine the number of mismatches segregated to the mother (Mm) or daughter (Dm) in a single division. Actual distributions are represented by gold (Dm) and purple (Mm) bars. Lines depict models of data: pink, single Poisson (P, k = 1); aqua, negative binomial (nb). c, d Correlations between Mm and Dm counts from actual data (green, n = 50) and simulations (n = 1000) under two different models. In (c) top panel depicts a cell with converging replication forks from two replicons with different mutation rates. Bottom panel shows the correlation of simulated Mm and Dm values (blue) drawn from the full negative binomial and their linear regression. In (d), top panel depicts two cells replicating DNA with different mutation rates. Bottom shows the correlation of simulated Mm and Dm values (orange) and their linear regression. e Correlation between the number of mismatches per chromosome segregated to Mother (Mm) or Daughter cells (Dm). green, observed counts; orange, simulated counts from model in (d). f Simulated index of dispersion of full replication error counts from small cohorts (n = 50) assuming the models from (c and d).
