PT - JOURNAL ARTICLE AU - Xia Shen AU - Xiao Wang AU - Zheng Ning AU - Yakov Tsepilov AU - Masoud Shirali TI - Simple multi-trait analysis identifies novel loci associated with growth and obesity measures AID - 10.1101/022269 DP - 2015 Jan 01 TA - bioRxiv PG - 022269 4099 - http://biorxiv.org/content/early/2015/07/08/022269.short 4100 - http://biorxiv.org/content/early/2015/07/08/022269.full AB - Anthropometric traits are of global clinical relevance as risk factors for a wide range of disease, including obesity1,2. Yet despite many hundreds of genetic variants having been associated with anthropometric measurements, these variants still explain little variation of the traits3,4. Joint-modeling of multiple anthropometric traits, has the potential to boost discovery power, but has not been applied to global-scale meta-analyses of genome-wide association studies (meta-GWAS). Here, we develop a simple method to perform multi-trait meta-GWAS using summary statistics reported in standard single-trait meta-GWAS and replicate the findings in an independent cohort. Using the summary statistics reported by the GIANT consortium meta-GWAS of 270,000 individuals5, we discovered 359 novel loci significantly associated with six anthropometric traits. The “overeating gene” GRM5 (P = 4.38×10−54) was the strongest novel locus6-8, and was independently replicated in the Generation Scotland cohort (n = 9,603, P = 4.42×10−3). The novel variants had an enriched rediscovery rate in the replication cohort. Our results provide new important insights into the biological mechanisms underlying anthropometric traits and emphasize the value of combining multiple correlated phenotypes in genomic studies. Our method has general applicability and can be applied as a secondary analysis of any standard GWAS or meta-GWAS with multiple traits.