Population genomic analyses with sambaR - barrettlab/2021-Genomics-bootcamp GitHub Wiki

Population genomic analyses with sambaR

de Jong MJ, de Jong FJ, Hoelzel AR, Janke A. 2021. SambaR: an R package for fast, easy and reproducible population‐genetic analyses of biallelic SNP datasets, Molecular Ecology Resources.

sambaR github page

sambaR manuscript

The following two commands are run in the unix terminal

## Convert vcf to bed/map
PGDSpider2-cli -inputfile imputed.vcf -outputfile out.ped -spid VCF2PED.spid

## convert to raw/bim
plink --file out --allow-extra-chr --make-bed --recode A --out out2

## This should create a .raw and .bim file
## Edit the individual and population names in the .raw file to whatever you would like

Download SambaR: https://github.com/mararie/SAMBAR/archive/refs/heads/master.zip

OR

install.packages("devtools")
library(devtools)
devtools::install_github("mararie/SAMBAR")

The following commands are run in R

### Load SambaR, if you downloaded the package as a .zip file

source("C:/Analysis/R_analysis/SambaR-master/SAMBAR_v1.06.txt")

### load dependencies
getpackages()

### Import data (the raw/bim files here called with prefix "out4", and a geofile.txt, which contains gps coordinates for each individual

importdata(inputprefix="out4",geofile="geofile.txt",allow_edit = TRUE)


### Required: filter your data
filterdata(indmiss=0.25,snpmiss=0.1,min_mac=2)

### Conduct population structure analyses (LEA, PCoA, etc). Add 'quickrun=FALSE' for additional analyses

findstructure(Kmax=6,add_legend=TRUE, legend_pos='bottomright',legend_cex=3, symbol_size=3)

### If you get a message about the number of snps, type the following:
findstructure(do_continue=TRUE, quickrun=FALSE)

### Create maps with pie charts (LEA q-values) and TESS maps
create_sambarmaps(K_max=6,radius_ratio=50)

### Calculate distance metrics
calcdistance(nchroms=NULL)

### Calculate diversity metrics

calcdiversity(nrsites=NULL,legend_cex=2.5)

### Inbreeding and relatedness

calckinship()

### Selection analysis for Fst outliers with PCAdapt

selectionanalyses(do_pcadapt=TRUE,do_outflank=FALSE,do_fsthet=TRUE,export='pdf',dopiechart = FALSE)