Estimating unsampled viral diversity
Branch: master
Clone or download
Latest commit 6a15bf1 Aug 23, 2017
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
analysis Repo reorganization Aug 23, 2017
deprecated Moving old R work to deprecated directory Aug 23, 2017
manuscript Update gitignore Aug 23, 2017
poster Reorganize repo. Aug 20, 2016
.gitignore Update gitignore Aug 23, 2017
README.md Repo reorganization Aug 23, 2017

README.md

Using the coalescent to detect missed infections in phylogenetic trees

Stephanie Stacy1,2, Allison Black1,3, Gytis Dudas1, Trevor Bedford1

1Vaccine and Infectious Disease Division, Fred Hutchinson Cancer Research Center, Seattle, WA, USA, 2Williams College, Williamstown, MA, USA, 3Department of Epidemiology, University of Washington, Seattle, WA, USA

Introduction

In viral outbreaks, some infections are likely to be missed. Finding as many cases as possible is an important foundation for developing better treatment and containing the spread of pathogens. This project is ultimately designed to highlight clades containing possible undersampled pockets.

To achieve this, we first calculate the proportion of data expected under each internal node in a given tree. We extend this method to target nodes and branches within the tree expected to be over or under sampled based upon z-scores of these nodal probabilities. For a given tree structure, conditional probabilities describing the chance of seeing a new, unknown sample coalescing under a given node in the tree are calculated based on effective population size, interval lengths (captured by sampling and coalescent events), and number of lineages within intervals.

Organization