Extracting biological information from DNA-biomolecule interactions using machine learning and statistics

DNA-biomolecule interactions such as protein-DNA and RNA-DNA interactions give rise to many important biological phenomena, ranging from DNA replication to gene regulation. Understanding the features that govern the success of these DNA-biomolecule interactions can thus aid in the understanding of fundamental biological processes and in the development of therapeutics. While the strength of these DNA-biomolecule interactions is often determined by a large number of degrees of freedom, the abundance of biological data from high-throughput sequencing experiments and the development of statistical learning algorithms have made it easier to understand these seemingly intractable biomolecular rules. This thesis represents our efforts to use large sequencing datasets in conjunction with machine learning and other statistical techniques to extract the most salient degrees of freedom in specific DNA-biomolecule interactions. Using neural networks and other learning algorithms to analyze transcription factor ChIP-seq datasets, we first explored the example of protein-DNA interactions due to the SOX10 transcription factor, and how mutations in the SOX10 DNA-binding domain alter the protein-DNA interactions genome-wide. Using multimodal neural networks and other statistical techniques with transcription factor ChIP-seq and RNA-seq datasets, we next investigated the role of protein-DNA interactions in the ETS transcription factors family, with an emphasis on the interactions between isoforms/paralogues of the protein GABP with the DNA in the TERT promoter. Finally, using various linear models and dimension reduction methods with amplicon sequencing and histone modification ChIP-seq datasets, we probed the RNA-DNA and protein-DNA interactions that give rise to efficient prime editors. The mechanisms discovered via these DNA-biomolecule interactions can provide insights into methods of treating various diseases such as cancers and other genetic-based disorders.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Physics

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2024

Author dc:creator

• Yuan, Jimmy Bo

Contributors dc:contributor

• Song, Jun S
• Dahmen, Karin A
• Aksimentiev, Aleksei
• Perez-Pinera, Pablo

• Machine Learning
• Statistics
• Biophysics
• Genomics
• Bioinformatics
• Protein-dna Interactions
• Genome Editing
• Tumor Biology

Statement dc:rights

• Copyright 2024 Jimmy Yuan

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/127314

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F127314