Back

Development of polyphenotypic scores to prioritize detection of G6PD rs1050828-T carriers in African and African American populations

Lu, T.; Stein, D.; Zhang, W.; Itan, Y.; Paterson, A. D.

2025-07-17 endocrinology
10.1101/2025.07.17.25331703 medRxiv
Show abstract

IntroductionThe G6PD missense variant rs1050828-T is common in African and African American populations. It lowers HbA1c levels independent of glycemia, increasing the risk of underdiagnosis or delayed diagnosis of diabetes. We aimed to develop polyphenotypic scores (PPS) using routinely collected phenotypes to identify likely carriers. Materials and MethodsUsing data from 31,083 African or African American participants in the All of Us Research Program (AoU), we developed sex-specific PPS through multi-stage variable selection. We validated the PPS in independent African or African American individuals from AoU (N = 8,846), BioMe Biobank (N = 8,839), and UK Biobank (N = 6,811), and evaluated their utility among individuals without diagnosed diabetes. ResultsThe PPS achieved an area under the receiver operating characteristic curve ranging from 0.7614 to 0.8686 in males and from 0.6033 to 0.6933 in females for identifying rs1050828-T carriers. Among individuals without diagnosed diabetes and with HbA1c <6.5%, PPS-based screening reduced the number needed to screen for identifying one carrier by 65-90% in males and 40-60% in females, compared to random screening. DiscussionThe PPS may help identify likely rs1050828-T carriers in African and African American populations, supporting prioritization for targeted genetic testing or alternative diagnostic approaches for diabetes.

Matching journals

The top 6 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.