Assessment of HCV infection stage as recent or chronic using multi-parameter analysis and machine learning

Hepatitis C virus (HCV) usually establishes chronic infection, which is often asymptomatic at the early stages of disease. Unfortunately, no diagnostic criteria that can distinguish between recent and chronic HCV infections are available. Error-prone replication of HCV causes each patient to host a heterogeneous population of genetically related HCV variants. Therefore, it is usually supposed that intra-host HCV heterogeneity gradually increases over the course of infection. However, due to the complex nature of the structural development of HCV populations inside hosts being influenced by selective sweeps and negative selection during chronic infection [3][2], the accuracy of simple metrics for the assessment of genetic heterogeneity is insufficient for HCV infection staging.