Table 1. Association of polymorphisms in circadian negative-feedback loop genes with overall survival in GC patients.

Gene	SNP	Predicted function	Best Fitting Model	Training set			Validation set		Pooled analysis
				HRa (95% CI)	P	Bootstrapb (P < 0.05)	HRa (95% CI)	P	HRa (95% CI)	P
CRY1	rs3809236	TFBS	Dominant	0.85 (0.64–1.09)	0.501		0.96 (0.75–1.21)	0.713	0.92 (0.69–1.28)	0.579
	rs1056560	miRNA	Additive	0.72 (0.58–0.88)	0.021	96	0.74 (0.46–0.90)	0.023	0.65 (0.34–0.87)	0.001
CRY2	rs6798	TFBS	Dominant	0.88 (0.67–1.16)	0.422		0.97 (0.57–1.61)	0.869	0.95 (0.64–1.42)	0.287
	rs2292910	miRNA	Dominant	0.95 (0.76–1.19)	0.652		1.12 (0.51–2.50)	0.725	1.07 (0.72–1.58)	0.695
PER1	rs2735611	Splicing	Dominant	0.87 (0.68–1.11)	0.275		0.92 (0.58–1.48)	0. 841	0.95 (0.65–1.57)	0.881
	rs3027178	Splicing	Dominant	1.72 (1.19–2.35)	0.001	100	1.54 (1.07–1.98)	0.034	1.71 (1.25–2.34)	<0.001
PER2	rs2304669	Splicing	Dominant	0.92 (0.70–1.20)	0.403		1.12 (0.83–1.58)	0.612	1.16 (0.70–1.52)	0.326
	rs934945	nsSNP	Dominant	0.76 (0.56–1.25)	0.482		0.91 (0.62–1.47)	0.862	0.89 (0.58–1.35)	0.933
PER3	rs228729	TFBS	Recessive	1.93 (1.31–2.85)	0.003	98	1.39 (0.82–1.97)	0.170	1.79 (1.29–2.93)	0.003
	rs228669	Splicing	Dominant	1.15 (0.95–1.40)	0.391		1.14 (0.79–1.42)	0.506	1.19 (0.86–1.51)	0.275
	rs2640908	Splicing	Recessive	1.44 (0.98–1.96)	0.076		1.31 (0.93–1.86)	0.127	1.27 (0.95–1.62)	0.113
	rs172933	TFBS	Recessive	1.21 (0.74–1.44)	0.203		1.17 (0.64–1.51)	0.320	1.25 (0.98–1.57)	0.062
	rs2859390	Splicing	Dominant	0.92 (0.72–1.18)	0.516		0.98 (0.73–1.31)	0.962	0.94 (0.69–1.43)	0.898

Note: The significant values were shown in boldface (P < 0.05).

HR indicates hazard ratio; CI, confidence interval; TFBS, transcription factor binding site.

^aAdjusted by age, sex, tumor site, tumor size, differentiation, TNM stage and chemotherapy where appropriate.

^bBootstrap analysis was performed using 100 replicates to determine statistics support.