Population attributable risk model

In this tutorial, we will use SeqSIMLA to simulate families with disease.

  • In this tutorial, we use Asian 500kb on chrom 1(download).
    -popfile ASN_500k.bed.gz
    -recfile ASN_500k.rec

  • In our example pedigree file(download), there are 20 families, with 1,380 people.
    We randomly choose two persons from each of the pedigrees in the pedigree file, then put them in a proband file(download).
    -famfile SAP.txt
    -proband probands.txt

    If you don't have a pedigree file, you can just use the option "default 3-generation families" to generate fixed 3-generation pedigrees.
    see "Output Options: -fam number" in User Manual.

  • One replicate of simulated data.
    -batch 1

  • Four options below are required in SeqSIMLA to generate the disease status by the population attributable risk model.
    The --mode-par tells SeqSIMLA to use the population attributable risk model.
    The -alpha 0.05 specifies the baseline penetrance. The disease prevalence in the general population would be close to 5%.
    We select sites 1, 200, and 3000 as the disease sites, --fixed-par means the par value is divided equally by the number of the disease sites, then the result are assumed as the odds ratio for the three sites.
    -par 0.05
    -alpha 0.05
Collect all file into a folder and placed in the same directory you run SeqSIMA

Execute the following command,
./SeqSIMLA -popfile data/ASN_500k.bed.gz -recfile data/ASN_500k.rec -famfile data/SAP.txt -proband data/probands.txt -folder test1 -header test -batch 1 -site 1,200,3000 --mode-par -par 0.05 -alpha 0.05 --fixed-par

With our example files, this simulation would take about 150 seconds.

Notice: If you don't want to make the command yourself, we provide a generate command user interface on our website.