Following Fig 2 in the main text, but for *N*_{0} = 1000 mtDNAs rather than *N*_{0} = 100. Normalised heteroplasmy variance *V*′(*h*) (left column) and copy number variance *V*(*N*) (right column) for mtDNA randomly distributed in networks. **A**: simulations for *h* = 0.1; **B**: simulations for *h* = 0.5; **C**: sum over state variables for *h* = 0.5; **D**: first-order Taylor expansion for *h* = 0.5. The three columns for each panel give decreasing network heterogeneity, expressed via different seed numbers, 4, 16 and 64 (more seed points give a more homogeneous network). In each panel, variances are given for different values of wild and mutant type network inclusion parameters *p* (horizontal axis) and *q* (vertical axis). White baseline reflects the null case from the analytic sum without any network inclusion.

(TIF)

Read more here: Source link