SMR was an efficient method to identify associations between gene expression and complex traits using summary data from GWAS and eQTL studies. One advantage of SMR was that it was useful to prioritize functionally relevant genes in a trait/disease associated locus. A major challenge of GWAS in interpreting which gene is functionally associated with the trait comes from the fact that the significant SNP represents a large region of LD. Regions of strong LD can be very large, and the significant SNPs have been found in perfect LD with the causal SNP hundreds of kilobases away [41]. The SMR analysis (including a SMR test and a HEIDI test) was useful to prioritize the functional gene in the associated region. The top associated GWAS SNP rs228769 was located in HDAC5 gene region, 60 kb downstream from ASB16-AS1 (Fig. 3). The gene HDAC5 may contribute to osteoporosis etiology by controlling sclerostin expression in osteocytes and regulating osteoblast differentiation [50]. Since there was no probe for the HDAC5 gene in both eQTL studies, the present study cannot exclude the causal role of HDAC5
