Predicting patient response to hepatitis B treatment through minimal covariance networks

Abstract: 

The poor response to some drugs used in antiviral therapy for a chronic hepatitis B infection may be linked to mutations in the surface antigen gene of the hepatitis B virus. Many of the mutations at positions within the surface antigen gene do not occur independently, and covarying pairs of these positions can be linked into a network. In this talk, I will discuss the formulation of binary integer programming problems for generating minimal covariance networks which identify the most biologically relevant positions that separate patient response groups, thereby indicating features associated with the failure or success of antiviral therapy.