prewas manuscript analysis

This repository contains the data and R scripts in support of the manuscript: prewas: Data pre-processing for more informative bacterial GWAS. The focus of this work is to assess the potential impact of variant pre-processing choices on bGWAS results.

Mansucript Authors

Katie Saund* (https://orcid.org/0000-0002-6214-6713), Zena Lapp* (https://orcid.org/0000-0003-4674-2176), Stephanie N. Thiede* (https://orcid.org/0000-0003-0173-4324), Ali Pirani (https://orcid.org/0000-0001-7810-0982), and Evan S. Snitkin (https://orcid.org/0000-0001-8409-278X)

*Equal contribution

Repository contents

This repository includes the R code necessary to perform analyses and generate the figures in the manuscript.

How to use this repository

data

The data direcotry contains three subdirectories:

• hpc
• local
• key

data/hpc contains data we generated using our high performance computing cluster (hpc). There are scripts available so that you could adapt them for your computer system, but they are slow and/or computationally intensive analyses. This directory contains some zipped files (.gz) that need to be unzipped for scripts in lib/local to run without error.

data/local contains the data the can be generated quickly on a desktop computer in R using the scripts in lib/local starting from data in data/hpc and data/key

Both data/local and data/hpc are subdivided by analysis.

data/key contains several files necessary for plotting the data correctly (color palettes and data labels).

lib

The lib directory contains two subdirectories:

• hpc
• local

lib/hpc contains example scripts and functions used to generate the data in data/hpc. These scripts will not run "as is." They are provided so that users could adapt the code to their particular computer system.

lib/local contains scripts to perform any data analysis necessary to convert data in data/hpc into a form ready to be plotted. The script lib/local/plot_figures.R will use the provided data in data/ to generate the figures in figures/. Scripts are written to be run from the prewas_manuscript_analysis/ directory.

Both lib/local and lib/hpc are subdivided by analysis.

figures

The plots found here were generated with the script lib/local/plot_figures.R The plots in this directory were finalized in Adobe Illustrator (joining panels together into one figure, resizing, etc..).

Sequence data

All genome sequences available on NCBI (see Table S1).

Contributors

Katie Saund, Zena Lapp, and Stephanie N. Thiede all contributed code to this repository.

Citation

prewas: Data pre-processing for more informative bacterial GWAS
Katie Saund, Zena Lapp, Stephanie N. Thiede, Ali Pirani, Evan S Snitkin
bioRxiv 2019.12.20.873158; doi: https://doi.org/10.1101/2019.12.20.873158