AI model predicts infectiousness of COVID-19 variants using machine learning and a risk assessment model. The model analyzes genetic sequences from 30 countries collected by GISAID. It detects 73% of variants causing at least 1,000 cases per 10 lakh people after one week, increasing to over 80% after two weeks. The strongest predictors of infectiousness are early trajectory of infections, spike mutations, and differences from the dominant variant. This approach can be extended to other respiratory viruses, infectious diseases.Can AI model detect fresh waves of COVID infection?
GeneralAI model predicts infectiousness of COVID-19 variants using machine learning and a risk assessment model. The model analyzes genetic sequences from 30 countries collected by GISAID. It detects 73% of variants causing at least 1,000 cases per 10 lakh people after one week, increasing to over 80% after two weeks. The strongest predictors of infectiousness are early trajectory of infections, spike mutations, and differences from the dominant variant. This approach can be extended to other respiratory viruses, infectious diseases.
Comments