Author

Yilin Ye

Date of Award

2024

First Advisor

Amanda Landi

Second Advisor

Michael Bergman

Abstract

This thesis explores the integration of FG-BERT and MolCLR, two advanced computational models, into a unified framework termed MolSynerNet, aimed at enhancing the predictive accuracy of molecular docking scores. The research addresses the challenge of accurately predicting molecular interactions essential for drug discovery, particularly focusing on allosteric EGFR kinase inhibitors—a critical target in cancer therapy. Through a systematic integration of sequence-based data (SMILES strings) processed by FG-BERT and structural data (molecular graphs) handled by MolCLR, the study seeks to leverage the complementary strengths of these models to improve prediction outcomes. The findings from extensive testing show that MolSynerNet significantly outperforms the baseline models—GIN for structural predictions and FG-BERT for sequence analysis—across several key performance metrics. Specifically, MolSynerNet achieved a prediction accuracy of 81.5%, reduced the root mean square error (RMSE) to 1.78, and increased the F1 score to 75.2, indicating a superior balance of precision and recall compared to the standalone models. These results validate the hypothesis that integrating different types of molecular data can lead to more accurate predictions of molecular interactions, thus enhancing the drug discovery process. The implications of this research extend beyond the immediate improvements in docking score predictions. By demonstrating the effectiveness of a hybrid computational approach, this work contributes to the broader field of computational drug design, suggesting that such integrative models can provide a more reliable and efficient pathway for the development of new therapeutic agents. This thesis not only advances our understanding of molecular docking mechanisms but also sets a foundation for future innovations in the integration of computational methods in drug discovery.

Simon's Rock Off-campus Download

Simon's Rock students and employees can log in from off-campus by clicking on the Off-campus Download button and entering their Simon's Rock username and password.

Share

COinS