Experimental and BF analysis of CO patterns in budding yeast.

Panels (A, B): Experimental System. (A) Spread yeast pachytene chromosomes fluorescently labeled for SC component Zip1, CO-correlated foci of ZMM protein Zip3, and terminally labeled at the end of Chromosome XV by a lacO/LacI-GFP array. (B) Positions of Zip3 foci along a single Chromosome XV bivalent were defined as shown. Panels (C–H): Experimental CO patterns for Chromosome XV. (C) CoC and ED relationships for a single representative Chromosome XV data set reflecting CO positions defined along >300 bivalents (as in (B)). Average CoC curve (black line) shows L_COC = 0.3 µm. (D–G) CoC curves and EDs for four independent experiments like that in (C). (D) shows the four individual average CoC curves; data set from panel (C) in black. (E) shows the four curves from four independent data sets and their average (in red). (F) shows the average of the four average CoC curves with the standard error at each inter-interval distance. (G) Shows the EDs for four independent experiments in (D–F) and the average (in red) with standard error. (H) Compares the average CoC curve and ED for an mlh1Δ mutant (blue) with those for WT (black; average of averages from panels (F) and (G)). Both ED relationships and CoC relationships in the mutant are as WT since Mlh1 acts very late (text). Panels (I–K): BF simulations of CO patterns for Chromosome XV (data from average of averages in panels (F) and (G). (I) Best-fit simulation (red) versus experimental data (black). Best-fit simulation specifies relatively even spacing of precursors (E = 0.6) and a constant number of precursors along the chromosome in all nuclei (B = 1). Other parameter values are in Table 1. (J, K). Experimental data and best-fit simulation data (black and red, from panel (I)) are compared with simulation using the same parameter values as the best-fit simulation except that precursors are either randomly spaced (Panel J; E = 0; green) or Poisson distributed among chromosomes in different nuclei (Panel K; B = 0; blue). Best-fit simulations for Chromosome IV and III data and for chromosome XIV in BR are shown in Figure S5 A–D; parameter values in Table 1. Importantly, even spacing is important for the best fit in all cases. In contrast, constant and Poisson distributions give very similar matches to experimental data except for the case of Chromosome III, where constant distribution must be required to ensure a sufficiently low number of zero-CO chromosomes (Figure S5 EF).