Table 2.

Quasi-Poisson Regression Fitted for Long (>10 cM) Runs of Homozygosity (NROH10; cM) for Individuals in the Global Sample

Predictor Variable	Estimate	Std. Error	t-Value	Pr(>|t|)
Intercept	0.95	1.34	0.71	4.80e-01
NROH1	−0.02	0.01	−3.13	1.76e-03
NROH2	0.06	0.00	17.22	4.62e-64
Heterozygosity	−6.56	3.27	−2.01	4.48e-02
Consanguinity score = 1	0.30	0.10	3.01	2.67e-03
Consanguinity score = 2	0.79	0.08	10.09	1.25e-23
Central Asia and Siberia	1.25	0.50	2.51	1.21e-02
East Asia	1.01	0.52	1.93	5.40e-02
Middle East and North Africa	0.70	0.53	1.30	1.93e-01
Oceania	−3.30	2.05	−1.61	1.08e-01
South Asia	1.10	0.53	2.09	3.70e-02
Sub-Saharan Africa	1.04	0.75	1.40	1.62e-01
Western Eurasia	0.39	0.53	0.74	4.57e-01

Abbreviations: NROH1, number of runs of homozygosity longer than 1 cM but shorter than 2 cM; NROH2, number of runs of homozygosity longer than 2 cM but shorter than 5 cM. Consanguinity scores of 1 and 2 represent populations that permit and prefer cousin unions, respectively. Regression (beta) coefficients for each region are expressed using NROH10 in the Americas as baseline. Ten genetic PCs (calculated for the full sample) were also included in the model, but their beta coefficients are not shown. Poisson dispersion parameter = 4.01, p-value = 0.