GGoutlieR

GGoutlieR (Geo-Genetic outlieR) is an R package for identifying and visualizing unusual geo-genetic patterns for landscape genomic studies.

Outliers with unusual geo-genetic association patterns can be identified through either genetics-based K-nearest neighbors (KNNs) or geography-based KNNs in the framework of GGoutlieR.

It provides a summary table with heuristic p-values allowing users to easily identify outliers from thousands of biological samples. Moreover, its visualization tool enables users to present unusual geo-genetic association patterns on a geographical map.

Details about GGoutlieR framework

To find the detailed algorithm of GGoutlieR, please check the supplementary material of our manuscript HERE (./paper/suppinfo.pdf).

Installation

You can install GGoutlieR either from CRAN or from this Github repository with the R commands below.

```

-----------------------

install from CRAN

-----------------------

install.packages("GGoutlieR")

-----------------------

install from GitHub

-----------------------

install dependencies

install.packages(c("stats4","FastKNN","foreach","doParallel","parallel","scales", "RColorBrewer","ggforce", "rlang", "tidyr", "utils", "rnaturalearth", "rnaturalearthdata", "sf", "ggplot2","cowplot"))

install GGoutlieR from Github

library(devtools) install_git("https://github.com/kjschmidlab/GGoutlieR.git") ```

Get started

You can find a more detailed tutorial with an example of GGoutlieR using a global barley landrace collection HERE (./vignettes/outlier_detection.pdf). The VCF file of the barley landrace collection is available HERE (./supp_data/IPKbarley_GBS1661landraces_LDpruned.vcf.gz).

Simple example

```

example of barley landrace data

library(GGoutlieR) data("ipkanccoef") # get ancestry coefficients data("ipkgeocoord") # get geographical coordinates

run ggoutlier (takes about 40 sec)

ggoutlierexample <- ggoutlier(geocoord = ipkgeocoord, gencoord = ipkanccoef, makefig = FALSE, pthres = 0.005, cpu = 4, klim = c(3,15), method = "composite", verbose = FALSE, multistages = FALSE) # switch off multi-stage test to reduce computational time

check the first few rows of the summary table

head(summaryggoutlier(ggoutlierexample))

visualize GGoutlieR results (set boundaries using plot_xlim and plot_ylim to focus on Eurasia and North Africa)

plotggoutlier(ggoutlierres = ggoutlierexample, gencoord = ipkanccoef, geocoord = ipkgeocoord, pthres = 0.005, maptype = "both", plotxlim = c(-20,140), plotylim = c(10,62), pierscale = 1.8, mapresolution = "medium") ```

Reference manual

The reference manual is at
https://cran.r-project.org/web/packages/GGoutlieR/GGoutlieR.pdf

Contributing

We appreciate your interest in using GGoutlieR in your study. You can contribute to the improvement of the package via following ways.

Questions and bug reports

Please open issues if you notice an issue with the GGoutlieR repository. You can also contact cheweichang92@gmail.com or karl.schmid@uni-hohenheim.de for your questions.

Pull request

If you would like to contribute to the code:

• Fork the GGoutlieR repository
• Contribute to your forked repository.
• Create a pull request.

Your changes or additions will be merged in the master branch of GGoutlieR repository if they pass required checks.

Preprint

Che-Wei Chang and Karl Schmid. 2023. GGoutlieR: an R package to identify and visualize unusual geo-genetic patterns of biological samples. bioRxiv. DOI: https://doi.org/10.1101/2023.04.06.535838