Table 3. Density of private SNPs in ENCODE regulatory regions of L-SSMD genes.

		Density of private SNP (per million bp)
Abbreviation	Description	Observed (#/Mb)	Control 1	Control 2	Control 3	Control 4
E	Predicted enhancer	2.07 (308/149)	1.54*	1.41*	1.76*	1.73*
TSS	Predicted promoter region including transcription start site	1.91 (408/214)	1.51*	1.23*	1.45*	1.82
CTCF	CTCF enriched element	1.89 (213/113)	1.71*	1.34*	1.56*	1.79
T	Predicted transcribed region	2.00 (4184/2092)	1.79*	1.52*	1.83	1.92
PF	Predicted promoter flanking region	1.79 (94/53)	1.46*	1.33*	1.69	1.96
R	Predicted repressed or low activity region	1.88 (10152/5400)	1.72*	1.45*	1.68*	1.79
WE	Predicted weak enhancer or open chromatin cis regulatory element	1.93 (102/53)	1.59*	1.63*	2.15	2.00
UNCL	Unclassified region	1.64 (833/508)	1.41*	0.84*	1.53	1.60

The symbol * indicates the SNP density in the corresponding control regions is significantly lower than that in the test regions of the outlier. The significance is assessed by one-tailed t test at the level of P = 0.001. Control 1: randomly selected non-outlier individuals to replace outlier individuals. Control 2: randomly selected genomic region that locate 10 Mb away from L-SSMD genes. Control 3: select randomly shuffled L-SSMD genes to the same amount of original gene set. Control 4: select randomly shuffled S-SSMD genes to the same amount of original gene set.