Allison Ebsen

Allison Ebsen

Allison Ebsen, a rising junior studying computer science at the University of Hawai’i at Mānoa, tested the robustness of a novel sequence alignment algorithm by assessing the impacts of changing score schemes on the outputs of the algorithm.

Double-strand breaks (DSB) to DNA can be lethal to cells if not repaired. The repair process can use microhomology (MH)—perfectly matching base pairs at the broken ends—to align the strands for repair. I came into BSRP with a lot of dreams, an open mind, and an uncertainty about what I needed to get out of my summer at the Broad; but I walked away with more than I could’ve hoped for. I am so thankful to everyone at the Broad who shared their time and gifts with me and my scientific journey. Every single day, I witnessed laulima (the practice of many hands working together in the spirit of aloha), and it was so empowering to experience all of us lifting each other up. As this summer journey at the Broad ends, it feels like I was a boat being rebuilt that is now being sent off back to sea. As bittersweet as it feels, boats were never meant to stay in the harbor. It was truly an honor and a blessing to be mentored at the Broad, and I am more excited than ever to continue on my scientific journey alongside everyone.However, improper repair can result in chromosomal structural variants (SVs), making the understanding of DSB repair mechanisms crucial for understanding maintenance of genome integrity. Existing SV calling algorithms identify microhomology but don’t explore sequences further from the breakpoint, which may play a role in the repair process. Our team developed a novel alignment algorithm that identifies statistically significant imperfect matching beyond the microhomology sequence. The algorithm analyzes these sequences by using a given scoring scheme, or point system, of rewards and penalties to calculate an alignment score for the sequences being compared. Using the algorithm and one score scheme, we tested this hypothesis by generating a background distribution of alignment scores from randomly generated SVs and compared the alignment scores of real SVs to the background. We hypothesized that the changes in score schemes would not significantly impact the outputs and findings of the algorithm. To analyze if there are differences, we compared relationships between different factors that may be impacted using data visualization techniques to compare the outputs across the 11 score schemes tested. We observed minimal differences between the outputs of our 11 score schemes, suggesting that changing score schemes doesn’t have a major effect on the algorithm’s outputs. This research will enhance our understanding of DSB repair.

 

Project: Assessing the Impacts of Score Schemes on a Novel Sequence Alignment Algorithm

Mentors: Simona Dalin and Sophie Webster, Cancer Program