PREST-plus identifies pedigree errors and cryptic relatedness in the GAW18 sample using genome-wide SNP data

BMC Proceedings

Table 4 Relationship testing results for clear outliers in Figure 1 identified by analysis 1, and in Figure 2 identified by analysis 2

FID1	IID1	FID2	IID2	null reltype		pvalue	plausible reltype		pvalue
The 7 outliers identified by analysis 1
3	T2DG0300174	3	T2DG0300175	1	full-sib	0	11	MZ-twins	N/A
4	T2DG0400281	4	T2DG0400282	1	full-sib	0	11	MZ-twins	N/A
4	T2DG0400265	4	T2DG0400266	2	half-sib	0	9	half-sib+first cousin	0.254
21	T2DG2100946	21	T2DG2100947	2	half-sib	0	9	half-sib+first cousin	0.432
21	T2DG2100952	21	T2DG2100966	4	avunuclar	0	6	unrelated	0.891
4	T2DG0400207	4	T2DG0400260	6	unrelated	0	2	half-sib	0.328
4	T2DG0400207	4	T2DG0400247	6	unrelated	0	5	first cousin	0.752
The 4 outliers identified by analysis 2
9	T2DG0901244	10	T2DG1000566	6	unrelated	0	8	half-first cousin	0.112
8	T2DG0800497	9	T2DG0901244	6	unrelated	0.007	8	half-first cousin	0.673
21	T2DG2100951	25	T2DG2501033	6	unrelated	0	5	first cousin	0.633
4	T2DG0400207	4	T2DG0400247	6	unrelated	0	5	first cousin	0.712

Empirical p-values are based on 25,000 simulated replicates, with genotype data simulated under a specified relationship type. The simulating relationship type can be the null relationship defined by the given pedigrees (i.e. the null reltype) or another relationship type (i.e. the plausible reltype) as listed in Table 1. The possible plausible relationship types are not unique and the table provides the one with the highest p-values. Small p-value for testing the null reltype suggests that the observed genotype data are not compatible with the null relationship defined by the given pedigree, whereas large p-value for testing the plausible reltype suggests that the observed genotype data are compatible with the proposed alternative.

ISSN: 1753-6561