Two other SNPs have recently been suggested to influence VT-risk, STXBP5 rs1039084 and VWF rs1063856 . These were not available in the ����in silico���� GWAS, but using QC imputed data in the whole set of 1,961 cases and 2,338 controls, the rare allele of the VWF rs1063856 was marginally associated with the risk of VT , an association consistent with that previously reported . Conversely, we did not observe any trend of association for the STXBP5 rs1039084 rare allele , even if this OR was of similar amplitude with that observed in the MEGA study . These two associations were previously observed in a meta-analysis of studies gathering about 5,000 cases and 5,000 controls, underlying the low power of our study to detect modest genetic effect as 957054-30-7 already discussed above. Large GWAS samples gathering at least ,20,000 patients would be required in order to detect genome-wide significant ORs of ,1.10 and, for the moment, we are far from reaching such sample size by contrast to international consortia on coronary artery disease . Another limitation of this work could be related to the selection of the GWAS subjects. Controls were part of a national GWAS sample of French buy MK-1775 healthy individuals that were not matched to VT cases, in particular for gender and sex. Nevertheless, all known or suspected VT-associated loci were identified in our work suggesting a rather modest influence of imperfect matching between cases and controls. Conversely, VT patients homozygous for the FV Leiden or FII 20210A mutation or with anti-thrombin, protein C or protein S deficiencies were not included in this work. It is very unlikely that the selection on FV Leiden homozygosity had affected our results as the F5 gene is among the four loci that reached genome-wide significance in our study. Note that the FII 20210 mutation was not available in the imputed reference datasets. However, one cannot exclude that the other exclusion criteria may have affect our power to identify novel VT-associated variants, in particular through a modulation of anti-thrombiin, protein C or protein S levels. It is nevertheless worthy of note that the PROCR locus that was found influencing the most protein C levels in the ARIC GWAS , was among the top 8 most significant VT-associated loci in our GWAS.