Table 1.
Optimization results derived using simulations of cell cycling
| Objective function | Free parameters | Optimization setup | Optimization result | |||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| P.D. first 3000 cells | P.D. gametes | Aggregate fitness | MZ length | MZ width | L2 cell cycle | L4, L4 + 1, L4 + 3 cell cycle | Single cell cycle length | Prod. constraints | Exp. geometry | Pre-meiotic arrest | Max. MZ cell number | Max. MZ length/width | Notes | Objective function optimum | Optimal MZ length | Optimal MZ width | α L4 | α L4 + 1 | ||
| Optimization | 1 | ✓ | ✓ | ✓ | ✓ | ✓ | 2000 | 2000 | 11.74 | 359 | 1 | 709 | 198 | |||||||
| 2 | ✓ | ✓ | ✓ | ✓ | 359 | 2000 | 13.94 | 1 | 359 | – | – | |||||||||
| 3 | ✓ | ✓ | ✓ | ✓ | 2000 | 2000 | 12.05 | 1523 | 1 | – | – | |||||||||
| 4 | ✓ | ✓ | ✓ | ✓ | 200 | 2000 | 17.20 | 1 | 200 | – | – | |||||||||
| 5 | ✓ | ✓ | ✓ | ✓ | ✓ | 2000 | 500 | 9.62 | 112 | 1 | 48 | 91 | ||||||||
| 6 | ✓ | ✓ | ✓ | ✓ | ✓ | 2000 | 30 | 9.65 | 18 | 12 | 3.50 | 1.36 | ||||||||
| 7 | ✓ | ✓ | ✓ | ✓ | 2000 | 2000 | 9.91 | 1 | 216 | – | – | |||||||||
| 8 | ✓ | ✓ | ✓ | ✓ | ✓ | – | – | 1 | 1.01 | – | – | 1.00 | 1.02 | |||||||
| 9 | ✓ | ✓ | ✓ | ✓ | ✓ | – | – | 1 | 1.00 | – | – | 1.00 | 1.01 | |||||||
| 10 | ✓ | ✓ | ✓ | ✓ | ✓ | 500 | 30 | 9.72 | 19 | 12 | 2.14 | 1.94 | ||||||||
| 11 | ✓ | ✓ | ✓ | ✓ | ✓ | – | – | 9.85 | – | – | 2.30 | 1.43 | ||||||||
| 12 | ✓ | ✓ | ✓ | ✓ | ✓ | – | – | 2 | 8.88 | – | – | 2.34 | 2.79 | |||||||
| 13 | ✓ | ✓ | ✓ | ✓ | ✓ | – | – | 3 | 8.85 | – | – | 0.63 | 0.71 | |||||||
Notes on optimizations:
1: Optimization 8 was run with deleterious mutation rate U = 0.03, and optimization 9 with U = 0.48
2: Optimization 12 assumes a mutation rate dependent on cell cycle speed
3: Optimization 13 assumes that distal cells preserve an immortal DNA strand
Each row shows results for one kind of problem, defined by the objective function to optimize (PD or pedigree depth; see main text and “Methods” for details), the parameters that are free to evolve within bounds during the optimization procedure (which can include MZ dimensions and distal cell cycle lengths at various stages), and other characteristics grouped under “Optimization setup.” Results shown are derived using a grid search (see “Methods”) α is the ratio of cell cycle lengths between the distal end of the DMMZ and the proximal end of the MMZ (slower distal cell cycle for α > 1). See Additional file 1: Table S1 for an extended version of this table, including credible intervals derived from Markov chain Monte Carlo (MCMC)