This page describes how to create random GWAS data for plink. A fake phenotype is generated for which significantly associated SNPs are present in the data.

= Required tools =
 * Gtool (http://www.well.ox.ac.uk/~cfreeman/software/gwas/gtool.html)
 * Hapgen2 (!https://mathgen.stats.ox.ac.uk/genetics_software/hapgen/hapgen2.html)
 * Plink (http://pngu.mgh.harvard.edu/~purcell/plink/)

= Required data =
Genotypes on which generated data is based. For instance: https://mathgen.stats.ox.ac.uk/impute/impute_v1.html#Using_IMPUTE_with_the_HapMap_Data

= Procedure =
For this analysis we used chr 22 of the hapmap data:

 * Genotype data: https://mathgen.stats.ox.ac.uk/wtccc-software/rel24_poly/haplotype+legend_files_CEU_r24.tgz
 * Recombination data: https://mathgen.stats.ox.ac.uk/wtccc-software/recombination_rates/genetic_map_b36_CEU.tgz

== Generate the artificial GWAS data ==
{{{
hapgen2 -h hapmap_r24_b36_fwd.consensus.qc.poly.chr22_ceu.phased -l chr22.ceu.r24.legend -m genetic_map_chr22_CEU_b36.txt -o test -dl 14431347 0 2 10 -n 2000 2000
}}}
== Merging case and control ==
{{{
gtool -M --g test.cases.gen test.controls.gen --s test.cases.sample test.controls.sample --og test.gen --os test.samples
}}}
== Convert to ped/map ==
{{{
gtool -G --g test.gen --s test.samples --ped test.ped --map test.map --phenotype pheno --chr 22 --snp
}}}
== Test with GWAS ==
{{{
plink --noweb --file test --assoc --maf 0.05 --hwe -0.001 --1 --allow-no-sex
}}}
plink.assoc should now contain the SNPs with p-values. There should be one really significant hit and server lower significant hits that are in LD with 'causal' SNP