Drug response prediction (DRP) remains a critical area of research in cancer treatment. We propose a novel architecture for pan-cancer drug response prediction that leverages graph-based representations of both cellular and pharmaceutical data. Our approach transforms protein-protein interaction networks to capture unique cancer mutations in each cell line, while simultaneously representing drugs through their molecular structures. Unlike traditional DRP models, our dual-graph approach enables us to exploit the inherent structural relationships in both cell line and drug data. By applying graph machine learning algorithms to these representations, we develop a robust model that predicts drug-specific half-maximal inhibitory concentration dosages for individual cell lines.

In this work, we introduce Text2Net, a text-based computer network simulation engine, that introduces a transformative approach to computer network education. Text2Net leverages state-of-the-art large language models (LLMs) and artificial intelligence (AI) text generation tools to interpret and transform plain-text descriptions of network topologies, configurations and functionalities, into dynamic and interactive simulations. This tool utilizes natural language processing (NLP) and large language models, specifically OpenAI’s ChatGPT-4T, to facilitate an intuitive interface that allows users (e.g. students, educators, and trainer/trainee engineers) to configure and interact with network simulations without the need for learning multiple simulation tools thus simplifying hands-on learning of networking. Text2Net addresses significant educational barriers by automating complex setup processes and making net- work learning faster, more accessible and engaging. This paper explores Text2Net’s development, focusing on its innovative use of AI, the system’s architecture, and its impact on network education. We provide a comprehensive evaluation of the system’s usability, performance, and educational value through qualitative analyses to demonstrate its effectiveness. Text2Net is a stepping stone to potentially revolutionize computer network education.

Molecular property prediction is pivotal in drug discovery, enabling accurate identification of potential compounds with desired characteristics. We introduce MolHyGAN, a novel hypergraph attention network model that captures higher-order molecular interactions for enhanced predictive capabilities. MolHyGAN represents molecules as hypergraphs, where nodes correspond to molecular substructures extracted using Extended Connectivity Fingerprints (ECFP) or k-mer sequences, and hyperedges represent the molecules themselves. This hypergraph-based representation allows the model to effectively capture complex molecular relationships that are often overlooked in traditional graph neural networks. Key innovations in MolHyGAN include attention mechanisms that prioritize important molecular substructures, stratified scaffold splitting for robust generalization, and comprehensive experiments across benchmark datasets (BACE, BBBP, ClinTox, and SIDER). By leveraging ECFP with varying radii (R = 2, 4, 6) and k-mer lengths (K = 3, 5, 7), MolHyGAN demonstrates significant improvements in AUC-ROC scores, achieving state-of-the-art results, particularly on balanced datasets. For example, MolHyGAN achieved an AUC-ROC of 0.9752 on the BACE dataset and 0.9939 on BBBP, outperforming existing methods. This study highlights MolHyGAN’s potential to transform molecular property prediction, particularly for applications in drug discovery and chemical research. Future extensions will explore hybrid approaches that integrate additional molecular fingerprints and embeddings to further enhance performance.