TABLE 2.

Results of the MCMC method for gene–gene–environment interactions affecting serum triglycerides

										Posterior probability
Risk set	APOA5a	LIPC2	LIPC1b	LIPC3b	SEX	SMOKE	GLU	N	Mean	General	Specific
1	HE/HO	HO				Y	IFG/IGT/T2D	5–9	2.73 (2.13–3.47)	0.24	0.24
2					M		IFG/IGT/T2D	65–630	1.5 (1.41–1.78)	0.86	0.69
3	HE/HO		WT/HO				T2D	16–22	3.03 (2.62–3.51)	0.16	0.13
4	HE/HO						T2D	18–26	2.85 (2.47–3.3)	0.68	0.58
5		HE/HO			M	Y	IGT/T2D	91–181	2 (1.82–2.16)	0.52	0.52
6	HE/HO		WT/HO			Y	IFG/IGT/T2D	35–74	2.07 (1.81–2.38)	0.45	0.45
7				WT/HO	M	Y	IGT/T2D	103–210	2 (1.82–2.19)	0.50	0.50
8	HE/HO				M			195–291	1.4 (1.31–1.5)	0.98	0.78
9					M	Y		266–1096	1.2 (1.16–1.32)	0.98	0.72
10	HE/HO						IFG/IGT/T2D	6–108	2.08 (1.7–2.9)	0.15	0.15
11							IFG/IGT/T2D	278–905	1.28 (1.2–1.44)	0.99	0.96
12						Y	T2D	12–119	1.79 (1.6–2.19)	0.20	0.16
13						Y		106–1067	1.08 (1.01–1.54)	1.00	0.91
14							IGT/T2D	48–178	1.34 (1.21–1.94)	0.66	0.54
15					M			93–1016	1.1 (0.99–1.49)	0.66	0.54

N, the number of individuals in the risk group; GLU, glucose tolerance status; WT, wild type; HE, heterozygote; HO, homozygote; M, man; Y, yes; IFG, impaired fasting glycemia; IGT, impaired glucose tolerance; T2D, screen-detected type 2 diabetes patients; Mean, the posterior estimate of the mean phenotype of the risk group. Posterior probabilities of the risk sets affecting serum triglyceride levels are shown. For explanation of the posterior probabilities see Table 1.

^aThe SNP parameters (HE or HO) and (HE) were grouped together because they are never sampled in the same iteration (very few homozygotes).

^bThe SNP parameters (WT or HO) and (WT) were likewise grouped together. The mean number of individuals not in a risk set is 1236. All combinations with an average general posterior probability of >0.1 are included.