GLOSSI: a method to assess the association of genetic loci-sets with complex diseases

BMC Bioinformatics

Table 1 Parameter specification in the simulated examples

Scenario	Number of causal SNPs	RR (OR)*	MAF^† of causal SNPs	Whether causal SNPs were 'genotyped'
1–3	1, 5 or 20	1.07 (1.10)	0.25	Yes
4–6	1, 5 or 20	1.34 (1.52)	0.05	Yes
7–9	1, 5 or 20	1.34 (1.50)	0.25	Yes
10–12	1, 5, or 20	1.61 (2.00)	0.05	Yes
13–15	1, 5, or 20	1.61 (1.94)	0.25	Yes
16–18	1 or 5	2.00 (2.67)	0.25	Yes
19–21	1, 5, or 17^‡	1.61 (~1.94)	~0.25	No

*RR (OR) = relative risks (odds ratio) when a loci carries two disease alleles, assuming an additive model; ^†MAF = minor allele frequency of the Utah samples in the HapMap project. ^‡Three out of the 20 original causal SNPs, generated under the case where MAF equals 0.25, are not in high LD with any SNP genotyped in the HapMap phase I/II project. Disease prevalence and crossover rate are fixed at 25% and 1.0 centiMorgan, respectively, in all simulations.

ISSN: 1471-2105