The present study may have some limitations. First, in the Westra et al. eQTL data [24], there were only 5967 probes so that it had incomplete and sparse genomic coverage of relevant genes, which may lead to the fact that only four positive findings using the SMR test. Some potential genes may thus be missed. Second, in both eQTL data, there is no probe for HDAC5. So SMR test did not yield any signal of HDAC5 but we cannot exclude the importance of the gene. Despite this, we found ASB16-AS1 was another putative novel functional gene for BMD at the locus. Third, the Gene Tissue Expression (GTEx) project which was an NIH funded effort planned to generate RNA-seq expression profiles from > 40 tissues in a large genotyped human cohort [61], did not collect bone tissue and primary bone cells. We thus cannot get eQTL data from bone related cells/tissue from GTEx for our analyses at present. Fourth, SMR may lose power compared to standard GWAS when the true biological mechanism is independent of gene expression [17].
