For each GWAS locus, we examined whether the lead/tag SNP correlated with any expression trait in GHS data and when a significant association was found, we checked whether the expression trait was significantly associated with the risk factor under consideration. This analysis revealed that very few GWAS results were compatible with an effect mediated by gene expression at the locus ( Table 4 and Table S2). There were, however, two exceptions: the first one concerned the LPL locus, where the minor allele of rs17489282 was associated with higher HDL-cholesterol (P = 5.91×10−5) and LPL expression (P = 2.18×10−6), while HDL-cholesterol and LPL expression were positively correlated (P = 6×10−4), consistent with an effect mediated by LPL; the second one concerned the association between the 1p13.3 locus and LDL-cholesterol. This locus encompasses three potential candidate genes, CELSR2, PSRC1 and SORT1, and it has been suggested that CELSR2 or SORT1 could be responsible for the reported associations of this locus with LDL [3], [24], [25]. In our data, the minor allele of rs629301 (a perfect tag of the lead SNP identified by
