The next frontier: AI, multi-omics, and predictive models for weight gain in HIV.

PURPOSE OF REVIEW: People with HIV (PWH) are increasingly susceptible to excess weight gain and obesity after initiation of antiretroviral therapy. However, there is substantial variation in individual weight gain that is difficult to predict with clinical factors alone. We review emerging methods in machine learning and multi-omics that address the biology and prediction of weight gain and weight-associated conditions, persistent challenges, and future directions. RECENT FINDINGS: PWH continue to have increasing burden of excess weight gain and obesity. Obesity has a complex pathophysiology with an interplay between biological and environmental factors that make weight gain prediction challenging. Few studies in PWH have used omics or machine learning to capture weight gain trajectories. These studies highlight the promise and challenges of leveraging multi-omics and machine learning for modeling weight gain. SUMMARY: Advances in machine learning and scalable multi-omics have accelerated research defining pathways of human health and disease. These tools may have an important role in defining the biology and predictors of weight gain in PWH, ultimately to improve risk stratification and identify targeted interventions to improve health outcomes in PWH.