Statistical analysis

Association analysis was performed using chi-square test of independence and using odds ratios (OR) and 95% confidence intervals (Mantel and Haenszel, 1959). We used R package Hardy-Weinberg (Graffelman, 2015) to test if genotype distributions were under Hardy Weinberg equilibrium. A chi-square test of independence was used to compare the distribution of rs2910164 genotypes among the several parameters (histopathological, subgroups defined by immunohistochemistry, tumour grade, and lymph node invasion). Bartlett’s test was applied to compare the age of cancer diagnosis in different genotypes (Bartlett, 1937). Analysis of the expression data from TCGA was performed using Kruskal-Wallis test followed by Dunn test (Kruskal and Wallis, 1952; Dunn, 1964). Spearman correlation was used to check the correlation of miR146a and the genes BRCA1, TRAF6, IRAK1, TRAF3IP1.

Alpha (α) was set at 0.05 for all tests and all analyses were performed with R Software (R core team, 2017) with the packages Nortest and readxl (Gross and Ligges, 2015; Wickham and Bryan, 2017).