lmoncla / ncov-WI-within-host

analysis of within-host diversity of SARS-CoV2 sampled in Wisconsin

ncov-WI-within-host

Analysis of within-host sequence data generated by Katarina Braun and Gage Moreno, from SARS-CoV2 patients sampled in Wisconsin.

The goal of this project is to characterize within-host diversity during acute SARS-CoV2 infection. Gage and Kat and have sequenced 142 samples collected from SARS-CoV2 infected individuals in Wisconsin, USA, including a set of 25 households for which samples from multiple individuals were taken. All 142 samples were sequenced in duplicate.

This repository contains scripts for reading in intersection vcfs (vcfs containing only variants that were identified in both biological sequencing replicates) and summarizing patterns of within-host variation. There are also scripts to query whether variants detected within-host are present in the global phylogeny and to quantify aspects of within-host variant transmission. JSON files for the Wisconsin-specific Nextstrain builds are being hosted by Gage Moreno at https://github.com/gagekmoreno/Wisconsin-SARS-CoV-2.

Individuals are welcome to borrow and use this code if it is helpful, just let me know what you are up to.

Install

This repo requires a baltic/baltic.py, which is available here. I have also placed a working version within this repo, and all scripts within this repo point to that local baltic version. Every notebook also requires R and ryp2, a package for using R within python jupyter notebooks.

Project Structure

• data/auspice-jsons/: contains JSON trees viewable with Nextstrain. These JSONs correspond to frequently-updated Wisconsin-specific Nextstrain builds curated by Gage Moreno and updated here
• data/vcfs-all/: contains tab-separated files for each sample analyzed in this dataset. These files only contain SNVs detected in both biological sequencing replicates for a given sample. These serve as the input files for all analyses run in this repository.
• data/consensus-sequences is a file containing all consensus sequences generated from Wisconsin and shown in the phylogenies.
• scripts: contains ipython notebooks used to analyze the vcf files and generate figures