Drug Repositioning: Sitagliptin as a Promising Anti-Hepatitis B Virus Drug Candidate In Silico, In Vitro, and In Vivo.

Hepatitis B virus (HBV) causes chronic infection in 250 million people worldwide. Chronic HBV carriers have a high risk of fibrosis, cirrhosis, and hepatocellular carcinoma. Drug resistance often develops during long-term antiviral treatment using HBV polymerase/reverse transcriptase inhibitors, making anti-HBV therapies as well as drug development more complicated. In this paper, a machine learning method based on large scale of gene expression profiles was used for anti-HBV drug screening. We acquired the HBV infection gene expression profiles and 1,703 chemical compounds perturbation gene expression profiles to calculate the corresponding gene expression signatures. The signatures were clustered via Non-negative matrix factorization (NMF) to generate a list of candidate drugs. Ultimately, ten candidate anti-HBV drugs were predicted and we retrieved them from the PubMed (https://www.ncbi.nlm.nih.gov/pubmed/) database to generate literature support about the reliability of the proposed method. In addition, we conducted cytotoxicity screening as well as in vitro and in vivo experiments to further validate the effectiveness of the candidate drugs. Based on the complete validation process, this rapid and cost-effective method was shown to be powerful for predicting anti-HBV drug candidates, this approach could be used for drug repositioning for other diseases. Moreover, the drug sitagliptin, which we predicted and validated, is already FDA-approved and might be immediately tested in HBV infected patients who require new therapies.