Table 2.

Composite-likelihood approximation of the population recombination parameter ρ

Region	Group (n)	Median ρ (Morgans kb−1)	95% Confidence Interval
Chr. 1	Bella Maria (15)	0.424	0.370–0.562
Chr. 5	Bella Maria (15)	0.549	0.400–0.647
Chr. 21	Bella Maria (15)	0.534	0.514–0.560
Chr. 26	Bella Maria (15)	0.357	0.338–0.392
Chr. 1	El Huayco (12)	0.004	0.004–0.004
Chr. 5	El Huayco (12)	0.002	0.001–0.004
Chr. 21	El Huayco (12)	0.002	0.001–0.003
Chr. 26	El Huayco (12)	0.005	0.002–0.016
Chr. 1	Ardanza (11)	0.005	0.005–0.005
Chr. 5	Ardanza (11)	0.003	0.002–0.003
Chr. 21	Ardanza (11)	0.002	0.000–0.004
Chr. 26	Ardanza (11)	0.002	0.001–0.002
Chr. 1, simulated	FSC_r (10)	78.886	77.023–80.739
Chr. 1, simulated	FSC_n (10)	0.001	0.000–0.007
Chr. 1, simulated	BS_n (10)	0.000	0.000–0.000

Positive approximations of ρ for T. cruzi I isolates from Bella Maria differ from estimates derived for synthetic non-recombinant controls. The FSC_n control represents ten 3.1 Mb chromosomes simulated without recombination in fastsimcoal2⁷⁷. The confidence interval around ρ estimates for FSC_n overlaps zero. It also overlaps estimates for El Huayco, Ardanza and BS_n, a second synthetic non-recombinant dataset generated by BAMSurgeon⁷⁶ simulation approach (see Methods). Results from chromosome simulation with the recombination rate r set to 3.2 × 10^–4 (FSC_r) demonstrate the sensitivity of the LDhat⁷⁴ interval program applied to 100,000 diploid individuals under a finite-sites model of evolution