An attention-driven framework for drug repurposing against human metapneumovirus: Integrating predictive modeling...

Area of Science:

• Virology
• Computational Biology
• Drug Discovery

Background:

• Human metapneumovirus (HMPV) causes significant respiratory illness, especially in vulnerable populations.
• No approved antiviral treatments currently exist for HMPV infections.
• Drug repurposing is a cost-effective strategy for identifying novel therapeutics.

Purpose of the Study:

• To develop and apply attention-based machine learning methods for predicting novel drug candidates against HMPV.
• To construct a dedicated dataset for identifying potential anti-HMPV therapeutics.
• To computationally evaluate promising drug candidates through molecular docking.

Main Methods:

• Development of a computational framework utilizing machine learning and deep learning, specifically attention-based architectures.
• Creation of a specialized dataset for training and validating predictive models.
• Inclusion of molecular docking studies to assess the binding potential of predicted drugs.

Main Results:

• Attention-based methods demonstrated promising predictive performance, particularly with larger datasets.
• Tilorone and oseltamivir were identified as potential drug candidates for HMPV.
• Docking studies supported the potential therapeutic significance of the identified compounds.

Conclusions:

• Attention-based deep learning offers a viable approach for computational drug repurposing against HMPV.
• Tilorone and oseltamivir warrant further laboratory investigation for HMPV treatment.
• This study highlights the potential of computational methods in accelerating the discovery of antivirals.