Table 2.

Most significant single-variant associations (p < 5 × 10⁻⁷) detected in the present GWAS.

SNP	REF allele	ALT allele	REF allele frequencya,b	Z-scoreb	p	Directionb	I2 (p)c	Position (chr:bp)	Position within gene	Gene symbol	Distance from gene (bp)
rs6035856	G	T	0.55	5.329	9.88 × 10−8	+++++++	0 (0.91)	20:2188542	intron	LOC388780	–
rs6035857	A	C	0.55	5.329	9.88 × 10−8	+++++++	0 (0.91)	20:2188544	intron	LOC388780	–
rs6082416	G	C	0.55	5.153	2.56 × 10−7	+++++++	0 (0.85)	20:2195832	downstream	LOC388780	2035
rs6137325	G	T	0.59	5.147	2.65 × 10−7	+++++++	0 (0.91)	20:2187943	exon	LOC388780	–
rs2094530	C	T	0.22	5.098	3.43 × 10−7	?--???-	0 (0.82)	13:51564457	downstream	GUCY1B2	4190
rs6047381	C	T	0.58	5.089	3.59 × 10−7	+++++++	0 (0.84)	20:2185357	upstream	LOC388780	2217
rs6137326	C	T	0.59	5.083	3.72 × 10−7	+++++++	0 (0.91)	20:2187944	exon	LOC388780	–
rs6132418	A	T	0.58	5.08	3.78 × 10−7	+++++++	0 (0.85)	20:2186281	upstream	LOC388780	1293

^aAverage allele frequency computed for reference (REF) alleles over all the datasets analyzed.

^bAllele frequencies, meta-analysis Z-scores, and directions of effect refer to REF alleles. Directions of effect are reported for each single dataset analyzed, in the following order: AGS, Finland, France, Holland, Hungary, ENall1, ENall2. “+” means that the major allele is the risk allele, while “?” indicates that the variant was not tested in the corresponding dataset.

^cI² test for heterogeneity of genetic effect across datasets (the closer to “0”, the more homogenous is the genetic effect).