Detecting autozygosity through runs of homozygosity: A comparison of three autozygosity detection algorithms

BMC Genomics

Table 2 Top Regression Power Results

Program	Power	Length	Est kb	Het	α	β	d'	LD-pruning
Autozygosity within 20 generations (low genotyping error)
PLINK	0.696	50 SNPs	1108	0	0.003	0.33	3.19	Moderate
PLINK	0.691	80 SNPs	919	0	0.003	0.31	3.2	Light
PLINK	0.670	55 SNPs	1219	0	0.002	0.45	3.02	Moderate
PLINK	0.662	80 SNPs	918	1	0.005	0.23	3.28	Light
Autozygosity within 20 generations (high genotyping error)
PLINK	0.657	45 SNPs	949	0	0.004	0.29	3.19	Moderate
PLINK	0.636	55 SNPs	1160	0	0.002	0.47	2.97	Moderate
PLINK	0.628	50 SNPs	1054	0	0.003	0.37	3.07	Moderate
PLINK	0.608	60 SNPs	1265	1	0.003	0.41	3.01	Moderate
Autozygosity within 50 generations (low genotyping error)
PLINK	0.725	65 SNPs	746	0	0.005	0.40	2.83	Light
PLINK	0.712	50 SNPs	574	0	0.01	0.23	3.06	Light
PLINK	0.710	60 SNPs	689	0	0.006	0.36	2.88	Light
PLINK	0.709	45 SNPs	517	0	0.012	0.18	3.14	Light
Autozygosity within 50 generations (high genotyping error)
PLINK	0.671	35 SNPs	738	0	0.007	0.38	2.76	Moderate
PLINK	0.652	40 SNPs	844	0	0.004	0.47	2.69	Moderate
PLINK	0.649	45 SNPs	949	0	0.003	0.55	2.64	Moderate
PLINK	0.649	80 SNPs	777	1	0.006	0.40	2.73	Light

Listed are the top four repression power results for detecting autozygosity within the past 20 and 50 generations, and within low and high SNP genotyping error. Low genotyping error = 0.2% genotyping error rate. High genotyping error = 1% genotyping error rate. Length = Minimum SNP or cM length to call a segment. Est kb = Minimum expected kb distance for the given length. Het = Number of Heterozygotes allowed in a segment. α= Type 1 error rate. β= Type 2 error rate. d'= d-prime estimate. LD-pruning = Level of LD-pruning. Moderate = Removal of SNPs within a 50 SNP window that had VIF greater than 2 (i.e. --indep 50 5 2 in PLINK). Light = Removal of SNPs within a 50 SNP window that had VIF greater than 10 (i.e. --indep 50 5 10 in PLINK).

ISSN: 1471-2164