An Attention-Driven Framework for Drug Repurposing Against Human Metapneumovirus: Integrating Predictive Modeling With Docking Validation Publisher Pubmed



Roudsari AM ; Hooshmand M ; Majidifar S ; Danaeifar M ; Rabiee HR
Source: Antiviral Research Published:2026

Abstract

Human metapneumovirus (HMPV) is a member of the Paramyxoviridae family of viruses and has been associated with significant morbidity and mortality in recent years. HMPV poses a particular threat to the health of the elderly, young children, and individuals with compromised immune systems, particularly affecting the respiratory system. Currently, there are no approved drugs for the treatment or prevention of HMPV. Given the high cost of developing new drugs, computational drug repurposing has become a crucial strategy in drug discovery, enabling the repurposing of existing drugs for new targets. This study introduces attention-based methods for predicting novel therapeutics against human metapneumovirus. To facilitate this research, a dedicated dataset was constructed to identify potential anti-HMPV drug candidates. The proposed framework employs machine learning and deep learning techniques, with an emphasis on attention-based architectures, to generate these predictions. The results from our attention-based methods are promising, especially when there is a larger number of samples available. Among the predicted drugs, tilorone and oseltamivir show particular promise for further laboratory confirmation and testing. Additionally, this work includes a docking study of the proposed drugs, reinforcing their potential significance in the treatment of HMPV. © 2026 Elsevier B.V.