Combining genomic data and infection estimates to characterize the complex dynamics of SARS-CoV-2 Omicron variants in the United States

SARS-CoV-2 Omicron surged as a variant of concern in late 2021. Subsequently, several distinct Omicron variants have appeared and overtaken each other. We combined variant frequencies and infection estimates from a nowcasting model for each US state to estimate variant-specific infections, attack rates, and effective reproduction numbers (R t ). BA.1 rapidly emerged, and we estimate that it infected 47.7% of the US population between late 2021 and early 2022 before it was replaced by BA.2. We estimate that BA.5, despite a slower takeoff than BA.1, infected 35.7% of the US population, persisting in circulation for nearly 6 months. Other Omicron variants - BA.2, BA.4, and XBB - together infected 30.7% of the US population. We found a positive correlation between the state-level BA.1 attack rate and social vulnerability and a negative correlation between the BA.1 and BA.2 attack rates. Our findings illustrate the complex interplay between viral evolution, population susceptibility, and social factors during the Omicron emergence in the US.