There has been an increasing amount of interest in modeling virus propagation in recent years. However, the group-based infection mechanism of computer viruses is not well understood and the selection of infection function in virus propagation modeling has not been well studied. This paper describes a point-to-group (P2G) infection mode to describe virus propagation in networks with information sharing groups. Simulations compare the constant infection and Ⅰ-type infection functions with the new E-type infection function in the small-world-network environment. The simulation results show that the E-type infection function shows superior performances to that of the traditional Ⅰ-type infection function in modeling the P2G virus infection mechanism and the Ⅰ-type infection function shows better performance in modeling the random infection mechanism.