Andover High School
Medical and Population Genetics
Service dogs play an extremely important role in policework, firefighting, disability healthcare, and other roles. In order to reduce training costs, it would be extremely useful to be able to algorithmically predict whether or not a dog is genetically fit to become a service dog using only genomic information. As a first step towards this long-term goal, Jolene wrote a machine learning code using the random forests algorithm to predict what type of white-spotting pattern a dog would exhibit given only its genetic data—a phenotype that is much simpler to predict than complex behavioral phenotypes like training viability. Jolene was able to show that classification algorithms could predict the white spotting pattern very well in some cases, and not as well in other cases. The work that Jolene did will be refined by her research group using other machine learning algorithms and other traits, until an algorithm is eventually developed that can predict training viability.
Jolene has always loved STEM: she has a history of both coding and designing robots. However, Jolene had never encountered the field of computational biology before this summer. “The BSSP seemed like an amazing opportunity that would expose me to computational biology in a research setting,” said Jolene. “Through BSSP, I discovered my interest in computational biology, and wish to pursue a degree in computational biology or bioinformatics. Additionally, I am now considering going into scientific research.”