Hepatitis B virus (HBV) infection is a worldwide health problem, with more than 1 million people died from liver cirrhosis and hepatocellular carcinoma (HCC) each year. HBV infection could result in the progression from normal to serious cirrhosis which is insidious and asymptomatic in most of the cases. The recent development of DNA microarray technology provides biomedical researchers with a molecular sight to observe thousands of genes simultaneously. How to efficiently extract useful information from these large-scale gene expression data is an important issue. Although there exist a number of interesting researches on this issue, they used to deploy some complicated statistical hypotheses. In this paper, we propose a multi-information-based methodology to score genes based on the microarray expressions. The concept of multi-information here is to combine different scoring functions in different tiers for analyzing gene expressions. The proposed methods can rank the genes according to the degree of relevance to the targeted diseases so as to form a precise prediction base. The experimental results show that our approach delivers accurate prediction through the assessment of QRT-PRC results.