Table 2.

Markers and genes with a possible association with the testicular dysgenesis syndrome (TDS) phenotype

Gene	Marker*	Selection method†	Phase	Risk allele	Control RAF‡	OR (95% CI)		punadj	padj	Optimal genetic model§
						Per allele	Optimal genetic model
KITLG	rs1352947	TGCC	Discovery	T	0.81	1.56 (1.19 to 2.05)	1.52 (1.18 to 1.97)	3.1×10−3	1	Additive
	C/T		Replication		0.82	2.11 (1.48 to 3.03)	1.93 (1.39 to 2.69)	8.5×10−5	2.0×10−3
TGFBR3	rs12082710	ISB	Discovery	T	0.58	1.35 (1.10 to 1.65)	1.77 (1.33 to 2.36)	2.4×10−4	1	Recessive
	T/C		Replication		0.59	1.27 (0.99 to 1.63)	1.52 (1.08 to 2.15)	1.6×10−2	3.8×10−1
BMP7	rs388286	Pathway	Discovery	C	0.47	1.34 (1.10 to 1.63)	1.36 (1.11 to 1.67)	2.3×10−3	1	Additive
	C/T		Replication		0.47	1.29 (1.01 to 1.66)	1.28 (1.01 to 1.62)	4.1×10−2	9.9×10−1
HOXDx	rs17198432	TDS	Discovery	A	0.07	2.31 (1.66 to 3.26)	2.58 (1.82 to 3.70)	4.7×10−8	2.2×10−2	Dominant
	C/A		Replication		0.11	0.96 (1.43 to 0.64)	0.97 (0.64 to 1.48)	8.7×10−1	1

^*The marker with the lowest p value in the discovery cohort, among the markers tagging a gene, is presented.

^†Four different approaches were used for the selection of markers: TDS, single-marker genome-wide association study (GWAS) on all TDS sub-phenotypes; TGCC, single-marker GWAS on the testicular germ cell cancer (TGCC) subset of cases; Pathway, aggregated effect in pathways; ISB, integrative systems biology by combing evidence of association from several data types.

^‡Risk allele frequency.

^§The genetic model with strongest association among the models tested by the MAX test (selected on discovery cohort).