Estimation of SARS-CoV-2 infection rates using phylogenetic summary statistics.

SARS-CoV-2 infection is confirmed via a nose or mouth swab test. While accurate for identifying individual cases, this testing method has underestimated the total number of SARS-CoV-2 cases, due in part to frequent symptomless infections which can go untested. Additionally, with inequitable access to testing, some communities have significantly inferior case reporting than others. Since SARS-COV-2 case counts inform public health measures, an accurate estimate benefits all communities. To date, researchers have attempted to estimate the actual number of SARS-CoV-2 infections using two methods. The first method utilizes antibody testing to compare the number of individuals with markers of SARS-COV-2 infection in their blood to case counts. Unfortunately, this method can be subject to sampling bias and is becoming ineffective with increasing vaccinations which results in antibody creation in the vaccine recipient. The second method fits models of pandemic spread to SARS-CoV-2 case count data, which can oversimplify virus spread and lead to inaccurate estimates. Seeking to enhance infection estimation accuracy, we will utilize the SARS-COV-2 evolutionary history reconstructed from genetic sequences that are already generated. Using computer simulations, our goal is to find a link between features of a tree which represents how infections are related to each other (i.e. the shape of the tree) and infection numbers. Subsequently we will develop a computer program to estimate SARS-COV-2 infection numbers based upon features of the tree. Enhanced accuracy of infection estimates will provide vital information to support public health measures in the ongoing SARS-COV-2 pandemic by providing a more accurate representation of the true number of infected individuals.