Graph Neural Network-Driven Interactive Visualization of Signal Pathways for Personalized Drug Target Discovery

Poster Number: P163
Presentation Time: 05:00 PM - 06:30 PM

Abstract Keywords: Bioinformatics, Cancer Prevention, Cancer Genetics, Computational Biology, Pharmacogenomics
Working Group: Genomics and Translational Bioinformatics Working Group
Primary Track: Applications
Programmatic Theme: Translational Bioinformatics

The advent of precision medicine has markedly transformed cancer treatment, shifting towards a model that recognizes and addresses the individual molecular differences among patients. This paradigm hinges on selecting therapeutic drugs aimed at specifically targeting the aberrant signaling pathways that fuel tumor growth and progression. However, the intricate nature of intracellular signaling networks, combined with the diversity of genetic alterations in cancer, significantly complicates the identification of optimal therapeutic targets. The web of interactions among genes and proteins within cancer signaling pathways, characterized by complex feedback loops and cross-talk, plays a crucial role in determining cellular fate. Alterations such as copy number variations (CNVs), insertions and deletions (INDELs), single nucleotide variations (SNVs), and gene expression changes can disrupt these pathways, propelling uncontrolled cell proliferation and survival. Addressing this issue necessitates a comprehensive understanding of these pathways and their interactions within the tumor microenvironment and the immune system.Utilizing Graph Neural Networks (GNNs) with advanced visualization tools offers a powerful approach to overcome challenges in personalized cancer treatment. This combination provides a clear platform for clinicians and researchers to understand and use model predictions for informed therapy choices. Our aim is to improve patient outcomes through targeted therapy by integrating these computational methods to navigate drug selection complexities. For each patient, we analyze DNA, CNV, and RNA expression data, aligning these with 15 cancer signaling pathways visualized to show genes, processes, and interactions clearly. This setup helps clarify pathway dynamics for better decision-making.

Speaker(s):
Hui Li, Phd
University of Texas Health Science Center at Houston

Author(s):
jinlian wang, PhD - UTHealth;