This pipeline aim to do genome wide scan of various measures of molecular population genomics. It suppose that you have vcf files containing 2 populations. This pipeline will do the following thing :

1. Formating vcf file in order to keep only allelic information
2. Phasing haplotypes
3. For each individual, randomly sample one haplotype
4. calculate for each window the statistics => ABCstat_loci.txt
5. create a simple manhatan plot for each statistics in => gw_plot.pdf

The following list of programs must be install on your machine:

• Snakemake
• Singularity

git clone https://github.com/EwenBurban/GWScan_2pop.git
cd GWScan_2pop
chmod a+x configure.sh
./configure.sh

NOTE that configure.sh require sudo privileges. If you cannot obtain admin permission, see https://cloud.sylabs.io/builder

Before launching the pipeline you must enter various information in 2 files :

• the config file (by default I call it config.yaml)
• the popfile

an exemple of those two file is aviable (exemple_config.yaml and exemple_popfile.csv)

The config file provide various information that are mandatory for the pipeline

• work_dir: #the complete path dir where your vcf files are stored
• binpath: #the complete path where GW_Scan is installed
• popfile: #the complete path of your the popfile
• locusLength: #window size of analyse
• nameA: #name of one species (must be the same name in popfile)
• nameB: # same here
• Nref: #this argument is a remainder of part of DILS script. You don’t need to put an exact value (by default I put 85000)
• mu: #this argument is a remainder of part of DILS script. You don’t need to put an exact value (by default I put 3e-8)
• rho_over_theta: #this argument is a remainder of part of DILS script. You don’t need to put an exact value (by default I put 1.0)

The popfile inform the pipeline about the population attribution of each individual that you choiced to put in your analyse. It’s wrote using the csv convention.

lpipe.sh is the script that launch the pipeline. Before launching you must :

• edit the binpath value (put the same dir than in the config file)
• put in comment the line that is you wont use depending of your cluster management system (with or without slurm)
  • NB : if you use slurm in your cluster, you must edit the cluster.json file

without slurm

lpipe.sh <config.yaml path>

with slurm

sbatch lpipe.sh <config.yaml path>