As described in the main text, we overcome all four obstacles by constructing PGIs that include both publicly available data and restricted data (including genome-wide summary statistics from 23andMe) and releasing them to the data providers, who in turn will make them available to researchers. We omit the participating dataset from the summary statistics used to construct that dataset’s PGIs, and we construct all the PGIs using a uniform methodology. We similarly construct principal components (PCs) of the genome-wide data using a uniform methodology and release them to the data providers to make available. Our methodology is described below in Sections II and III. The code we used for constructing the PGIs and PCs is posted on the SSGAC website (https://www.thessgac.org/pgi-repository).
