Electrical and Computer Engineering ETDs
Publication Date
Fall 11-5-2024
Abstract
This thesis investigates knowledge graphs with particular reference to the NIH-funded Common Fund Data Ecosystem Data Distillery project. By combining data from nine Common Fund projects and other sources, this project has created a large knowledge graph using Neo4j. To find new and undiscovered drug targets, UNM’s Illuminating the Druggable Genome (IDG) Data Coordinating Center has supplied data and use cases. Condensed Knowledge Graph, a condensed version that is based on the Data Distillery Knowledge Graph, improves usability for IDG applications. Condensed Knowledge Graph research endeavors to enhance data organization through the categorization of disease terms, examination of Cerebellar Stroke to evaluate its representation, and use of the knowledge Graph to Machine Learning (KG2ML) model to forecast relationships between genes, proteins, and diseases. This methodology unveils relationships that were previously unknown, demonstrating Condensed Knowledge Graph’s ability to offer more profound comprehension of intricate biological processes and progress our knowledge of disease mechanisms, eventually contributing to the discovery of new therapeutic targets and treatments.
Keywords
Knowledge Graph, Knowledge Graph to Machine Learning, Neo4j, Genomic analyis, Gene-Disease association
Document Type
Thesis
Language
English
Degree Name
Computer Engineering
Level of Degree
Masters
Department Name
Electrical and Computer Engineering
First Committee Member (Chair)
Dr. Balasubramaniam Santhanam
Second Committee Member
Dr. Jeremy J Yang
Third Committee Member
Dr. Christophe G Lambert
Recommended Citation
Tenkila Purushotham, Swastika. "Advancing Drug Discovery and Disease Understanding through Knowledge Graphs and Machine Learning Techniques." (2024). https://digitalrepository.unm.edu/ece_etds/692
