Advisor(s)
Shahid Mukhtar
Committee Member(s)
Song Li
John McDowell
Karolina M Mukhtar
Cranos Williams
Document Type
Dissertation
Date of Award
2023
Degree Name by School
Doctor of Philosophy (PhD) College of Arts and Sciences
Abstract
Studying plant gene function is essential for understanding the underlying mechanisms that govern plant growth, development, and environmental adaptation. This knowledge can improve crop yields and food security by developing plants resistant to diseases, pests, and environmental stress. Moreover, understanding plant gene function can also aid in environmental conservation efforts by providing insight into the mechanisms that plants use to adapt to changing environments and can help preserve endan-gered plant species. For several reasons, studying gene function in plants can be more challenging than in animals. Plant genomes are larger and more complex than animal genomes, making it harder to identify and study specific genes. Many plants have multiple copies of each chromosome, known as polyploidy, which can make it difficult to determine the function of specific genes. Few genetic tools and techniques are available for studying plant gene function than animal gene function. Additionally, plants are sessile organisms, meaning they are anchored in one place and cannot move in response to environmental changes. We can use several experimental techniques to study plant gene function, including genetic mapping, Tilling, CRISPR-Cas9, RNA interference (RNAi), transgenic plants, proteomics, and metabolomics. Though, experimental techniques for studying plant gene function have several disadvantages, including being time-consuming and re-source-intensive, having limited applicability, and having unintended effects. Along with experimental techniques, several computational techniques are also available to infer plant gene function, including Homology-based prediction, Gene ontology analysis, Co-expression analysis, Pathway analysis, Protein-protein interaction analysis, Machine learning, and Network biology. These computational techniques can complement experimental techniques and provide valuable insights into the function of plant genes, particularly for genes with unknown functions. In my research, I have developed computational tools and pipelines based on graph theory, such as a Graph convolution network to infer gene function (SeqGraphInfer) and integrate multi-omics graph data to identify critical element gene regulatory net-works (POTFUL and CentralityCosDist). Furthermore, I have developed a natural language processing workflow (PATHAK) to mine genes and their potential ontologies from research articles.
Recommended Citation
Kumar, Nilesh, "Enhancing Plant Functional Annotation By Integrating Graph Convolutional Networks and Natural Language Processing" (2023). All ETDs from UAB. 25.
https://digitalcommons.library.uab.edu/etd-collection/25
Comments
etdadmin_upload_962887