Browsing by Author "MUHAMMAD ALI RAZA"

Now showing 1 - 1 of 1
  • Thumbnail Image
    Item
    Prediction of Antiviral Peptides using Long-Short Term-Memory (LSTM) Based Convolution Neural Networks
    (UMT, Lahore, 2023) MUHAMMAD ALI RAZA
    The emergence of viral infections as a significant global health concern has propelled research into innovative therapeutic strategies, including the utilization of antiviral peptides. This thesis explores the efficacy of antiviral peptides in combatting viral infections and employs machine learning techniques, including Long Short-Term Memory (LSTM) networks and various other algorithms, to enhance our understanding of their antiviral potential. The first phase of this research involved a comprehensive review of antiviral peptides, highlighting their mechanisms of action, structural features, and their demonstrated effectiveness against a wide array of viral pathogens. The unique properties of antiviral peptides, such as their ability to target conserved regions of viral proteins, make them promising candidates for the development of novel antiviral therapies. In pursuit of optimizing the prediction of antiviral peptide candidates, an extensive dataset was compiled, containing sequences of known antiviral peptides along with non-antiviral peptides. This dataset was then employed to train machine learning models. Various feature extraction techniques were employed to convert peptide sequences into a format suitable for machine learning algorithms. The application of these techniques enabled the development of models that accurately discriminate between antiviral and non-antiviral peptides.