Comparison of the top 3 axes of variation (PCs) among patients with ALL of different self-reported races/ethnicities and among different reference populations (CEU, N=60; YRI, N=60; CHB/JPT, N=90 samples from the HapMap and 105 Native American samples). Descriptions on X-axes represent self-declared designations. Boxes include data between the 25th and the 75th percentiles. Note that the first (PC1, Panel A), second (PC2, Panel B), and third axis (PC3, Panel C) reflect genetic variation characteristic of African, East Asian, and Native American ancestry, respectively. Also note that self-reported Asians consisted of both South Asians and East Asians (top and bottom cluster of far right bar in Panel B, respectively), with varying levels of East Asian genetic ancestry.

A. Genetic ancestral composition of 2,534 children with ALL. Each patient’s ancestry is depicted as a column, whereas color represents the proportion of ancestry estimated for that patient (European, red; African, gray; Asian, green; Native American [NA], blue). Genetic ancestry was estimated using STRUCTURE. Patients were clustered using Ward clustering method, based on dissimilarity in genetic ancestry measured by 1-minus pair-wise correlation (see Supplementary Note). Higher levels of NA ancestry were linked to increased risk of relapse in all patients (B), and within the self-reported whites (C), and for those who did not receive delayed intensification (D), but not within those who did receive delayed intensification in the COG P9904/9905 trial (E). Although cumulative incidence of relapse is plotted separately for patients with <10% (red) vs ≥10% (blue) NA ancestry, all P values were estimated using Fine and Gray’s cumulative incidence hazard regression model treating NA ancestry as a continuous variable (see Supplementary Note for details on NA ancestry dichotomization).