Hybridsim: Simulator for generating allele data in isolated populations

Petteri Hintsanen

We propose a novel two-phase population simulation procedure
for generating diploid marker allele data in isolated
populations.  First, founder population is simulated using
an approximated coalescence process and a stepwise mutation
model in order to establish a realistic genealogy and
mutation history for founders.  In the second phase, a
separate genealogy and exact recombination history is
generated for the isolate with a discrete-generation
forward-in-time simulator.  This procedure enables an
accurate simulation of a large population isolate.

Any disease model can be specified into the population
afterwards, and finally a sample of desired type can be
ascertained for further study.  Any number of SNP and/or STR
markers from the coalescence process can be used as markers
in the final allele data.

We discuss also some implementation issues and describe our
own implementation (available for download), which supports
a dominant mutation disease model and various sampling
methods.