The Population Architecture using Genomics and Epidemiology (PAGE) study, to which I contribute, is overcoming some of the technical challenges of multi-ethnic genomic analyses (see A. B. Popejoy and S. M. Fullerton Nature 538, 161–164; 2016). It is yielding findings that are unattainable for homogeneous populations.

PAGE, funded by the US National Institutes of Health, uses genome-wide analyses of 50,000 phenotyped participants of mainly Hispanic and African ancestry. We aggregate results across several studies: for example, two are multi-ethnic, one involves only women and one is designed to address under-representation of Hispanic people (see go.nature.com/2hity2j).

We collaborated with other initiatives, including the Consortium on Asthma among African-ancestry Populations in the Americas, to develop an unbiased genotyping array for use across all major continental populations (www.pagestudy.org/mega). Statistical tools such as SUGEN (D. Y. Lin et al. Am. J. Hum. Genet. 95, 675–688; 2014) and GENESIS (M. P. Conomos et al. Am. J. Hum. Genet. 98, 127–148; 2016) can account for the admixed ancestry of individuals (a significant factor in almost every US minority), and for cohorts that include many ancestries.

Leveraging sample diversity in these and other ways has maximized the power of our genetic analyses.