Development and application of machine learning tools for the advancement of energy crop genetic manipulation using the CRISPR/Cas technology

Crop genomes, often highly complex and polyploid, are tailored to optimize specific traits by generations of breeding. Editing such genomes without disrupting decades worth of selective breeding is a challenging task. The CRISPR/Cas9 technology's ability to perform targeted mutations has made it a prime candidate for this purpose. The majority of software tools currently available for designing CRISPR/Cas9 guide sequences were developed for mammalian genomes, and focus specifically on the genome's coding regions. We have developed CROPSR | an open-source tool to design and evaluate gRNA sequences for CRISPR experiments. Advantages over competing tools include a genome-wide approach, which facilitates the design of CRISPR/Cas9 guide RNAs in non-coding regions of the genome, and accounts for polyploidy present in many crop genomes. Novel algorithms were developed to increase performance when evaluating crop genomes, but without sacrificing performance when evaluating mammalian genomes. Incorporating biophysical properties of the CRISPR nucleoprotein complex, particularly the secondary structure of the gRNA molecule, enabled a two-fold increase in the correlation between the predicted and experimental metrics for mutation efficiency. Additionally, software performance was optimized to reduce resource allocation and run time. The results presented in this work are the initial steps in a longer journey to improve the tools available editing crop genomes. CROPSR enables a significant reduction in the time needed to design a CRISPR/Cas9 mutation experiments in complex, polyploid genomes. The addition of biophysical characteristics of the CRISPR/Cas9 complex were considered to develop a novel approach for mutation performance evaluation, resulting in significant increases in predictive accuracy and enabling improving confidence in the generated guide sequences.

Name thesis:degree_name

Ph.D.

Level thesis:degree_level

Dissertation

Discipline thesis:degree_discipline

Informatics

Grantor

University of Illinois at Urbana-Champaign

Year dc:date

2022

Author dc:creator

• Muller Paul, Hans

Contributors dc:contributor

• Hudson, Mathew
• Moose, Stephen
• Marshall-Colon, Amy
• Zhao, Sihai

• Machine Learning
• Crop Sciences
• Crispr
• Cas9
• Sgrna
• Algorithm
• Genome-wide
• Software Tool

Statement dc:rights

• Copyright 2022 Hans Müller Paul

Language dc:language

en, eng

Handle dc:identifier

https://hdl.handle.net/2142/117566

OAI record oai:identifier

https://www.ideals.illinois.edu/oai-pmh?verb=GetRecord&metadataPrefix=etdms&identifier=oai%3Awww.ideals.illinois.edu%3A2142%2F117566