Degree
Doctor of Philosophy (PhD)
Department
Division of Electrical & Computer Engineering
Document Type
Dissertation
Abstract
Predictive modeling has revolutionized computational biology and molecular bioinformatics, enabling significant advancements in cancer therapy and functional enzyme characterization. Despite considerable progress, significant challenges remain in accurately predicting combinational cancer therapies and systematically representing enzyme functions for computational applications. Traditional methods struggle with capturing the complex interactions between drugs and biological networks, as well as representing hierarchical relationships within enzyme classifications. This dissertation addresses these limitations by developing advanced deep learning models tailored to enhance predictive performance in both domains.
First, a data augmentation strategy is introduced to improve anticancer drug synergy prediction by generating pharmacologically relevant drug pairs based on a new similarity metric. This approach enriches training datasets, mitigates overfitting, and enhances model robustness. Additionally, SynerGNet, a graph neural network, is developed to predict drug synergy by integrating heterogeneous biological features onto a protein–protein interaction network, achieving superior accuracy and providing insights into augmented data integration and model generalization. In parallel, this dissertation introduces EC2Vec, a deep learning-based encoding method that effectively captures the hierarchical and functional relationships of Enzyme Commission (EC) numbers. By leveraging structured embeddings and convolutional layers, EC2Vec enhances the computational representation of enzyme function, demonstrating its broad applicability in bioinformatics.
Overall, the research results in this dissertation demonstrate methodological advancements in predictive modeling for cancer therapy and enzyme function analysis. By improving the accuracy and applicability of deep learning models, this work contributes to the development of more effective, personalized treatment strategies. Furthermore, the findings offer valuable insights for future applications in metabolic engineering, drug discovery, and synthetic biology, paving the way for innovative solutions in biomedical and biotechnological fields.
Date
3-25-2025
Recommended Citation
Liu, Mengmeng, "Deep Learning Applications for Predictive Modeling in Cancer Therapy and Enzyme Encoding" (2025). LSU Doctoral Dissertations. 6700.
https://repository.lsu.edu/gradschool_dissertations/6700
Committee Chair
J. Ram Ramanujam
