Nat. Genet. 42, 759–763 (2010); published online 22 August 2010; corrected after print 27 August 2014

In contrast to the version of this article initially published, the authors now find no evidence to support association with esophageal squamous cell carcinoma susceptibility for rs13042395[T] at 20p13 in their original data, in two independent sets of cases and controls collected in other Chinese populations or in the joint analysis of these three studies.

The original signal at the C20orf54 locus (rs13042395) in the published Henan genome-wide association study (GWAS) was based on the scan of a total of 2,810 subjects (1,077 cases and 1,733 controls) with a per-allele odds ratio (OR) of 0.66 (95% confidence interval (CI) = 0.58–0.74; P =1.51 × 10−11), and P < 5 × 10−6 was reported for a genetically matched subset of 937 cases and 692 controls.

A new analysis by the authors (see Supplementary Table 8 in Nat. Genet. 46, 1001–1006, 2014) included a distinct genetically matched subset of the original genotypes (1,789 total subjects: 1,076 cases and 713 controls; OR = 0.80 (95% CI = 0.65–0.98); P = 0.03) and subjects scanned after publication of the initial Henan data (669 new subjects: 229 cases and 370 controls; OR = 1.06 (95% CI = 0.82–1.38); P = 0.63). The combination of the genetically matched and new subjects (n = 2,458 subjects in total: 1,375 cases and 1,083 controls) yielded OR = 0.88 (95% CI = 0.76–1.02; P = 9.12 × 10−2).

Thus, the original Henan scan is the only result that showed a significant association for rs13042395. None of the three new analyses performed confirmed a significant association for rs13042395. Moreover, this SNP was not significant in the National Cancer Institute GWAS (OR = 0.95, 95% CI = 0.86–1.05; P = 3.04 × 10−1), Beijing GWAS (OR = 1.07, 95% CI = 0.93–1.23; P = 3.34 × 10−1) or combined data (OR = 0.96, 95% CI = 0.90–1.03; P = 3.02 × 10−1; see Supplementary Table 8 in Nat. Genet. 46, 1001–1006, 2014).

In summary, the published association for rs13042395 did not replicate in additional analyses of data from Henan. Nor did it replicate in a GWAS from a similar high-risk population (National Cancer Institute) or in a GWAS from a low-risk population (Beijing). Thus, the original finding was likely the result of inadequate control for population stratification using the genetically unmatched subjects or, less likely, could have been due to chance alone.

The error has been corrected in the HTML and PDF versions of the article. In addition, Supplementary Figures 1–4 and Supplementary Table 1 have been corrected.