Study of DNA Physics by a High-Throughput Genome Mapping Technique

Abstract

Next generation sequencing (NGS) is a powerful tool to sequence DNA with a low error rate at an affordable cost. While NGS is successful with these features, its short-read length leads to difficulty in detecting structural variation of chromosome at a large scale. Labeling barcodes can help resolve these drawbacks by generating scaffolds and can detect structural variation in combination with NGS. Though the labeling barcode technique is well-developed experimentally, how interpreting the data correctly still needs improvement. The properties and behavior of DNA in confinement are not fully understood and thus there is disagreement between experiments and theory. The aim of this dissertation research is to improve data interpretation by correcting assumptions which were misused or even not considered in previous studies. Using a high-throughput genome mapping technique, we first found a strong correlation between the persistence length of long DNA and sequence composition, which was oversimplified in previous works. We developed a statistical terpolymer model to rationalize the experimental results. We next revisited a widely used DNA model, a neutral wormlike chain model, through a comparison of experimental data sets with theory. We did a dimensional analysis and proposed that DNA-wall electrostatic interactions, which were neglected in the model, are the cause for the disagreement between experiments and theory. We hope all the findings in this dissertation provide a deeper understanding of properties and behavior of DNA in confinement and inspire scientists toward a new research direction to make a more robust DNA model for data interpretation of DNA-based experiments.