Currently, we are amid a global pandemic caused by the SARS-CoV-2 viruses. This is the third, and most widespread, highly pathogenic novel coronavirus to emerge in the last 20 years. Upon emergence of these virus, there has never been a tool to help predict the affect these viruses would have. Countries are forced into reaction after the virus has begun to spread. In this project, we develop computational tools to analyze genomic data from coronaviruses and predict their pathogenicity. Our best model was only marginally better than random guessing, achieving an accuracy of 0.68 and MCC of 0.34 after leave one group out cross validation. Future work may be able to improve upon this method by using a Hidden Markov Model to correct predictions.

• This report represents the work of one or more WPI undergraduate students submitted to the faculty as evidence of completion of a degree requirement. WPI routinely publishes these reports on its website without editorial or peer review.

Creator

• Simpson, Parker

Subject

• Computing
• Health

Publisher

• Worcester Polytechnic Institute

Identifier

• E-project-081621-153814
• 27361

Keyword

• Bioinformatics
• Machine Learning

Advisor

• Korkin, Dmitry

Year

• 2021

Date created

• 2021-08-16

Resource type

• Major Qualifying Project

Major

• Bioinformatics & Computational Biology

Rights statement

• In Copyright