Snakemake workflow to run ELAI with different parameters and SNP sets, and scripts to get average of multiple runs

Scripts here are adapted on IFB cluster. Branch 'itrop' is for scripts adapted on iTrop.

Run ELAI on simulated genotypes for testing (1 chromosome only)

Input

• source files, test file, and snp file in ELAI input format containing all SNPs
• define SNP sets

Here is my config file.

Steps

• generate or spit SNP sets as predefined
• run ELAI for each set of parameters and snp sets
• evaluate the results by custom scripts

Here is my script to run Snakemake.

Run ELAI on real data (multiple chromosomes possibly)

Input

• vcf file containing all the source individuals
• vcf file containing all the test individuals
• define SNP sets

Here is my config file.

Steps

• spit the source and test genotypes by chromosomes
• simulate source populations using the source individuals and generate source files for ELAI
• generate test file for ELAI from the vcf file
• generate or spit SNP sets as predefined
• run ELAI for each set of parameters and snp sets
• rerun ELAI by snakemake --forcerun elai
• average results from ELAI runs by custom script
• merge results from runs of different snp sets by custom script

Here is my script to run Snakemake.

Some notes

• the workflow runs on singularity container: shub://vibaotram/singularity-container:myr_4-0-2_rstudio_1.3.sif
• may need to change R lib path in the script files, and install all the packages before running to prevent failure of parallel jobs