Computational resources for simulating under a spatial coalescent model
across heterogeneous landscapes and testing hypotheses about the
geography of genetic variation: QUETZAL-EGGS, -CRUMBS, -NEST and DECRYPT
Abstract
Spatially explicit coalescent models in which the underlying demographic
parameters are informed by the environment (either past, present, or
temporally and spatially changing environments) provide a framework for
hypothesis testing that incorporates geographic information about
genetically sampled individuals. This general approach - Integrated
Distributional, Demographic and Coalescent (iDDC) modelling - can be
used to explain how heterogeneous, dynamic landscapes shape the history
and genetic patterns of a species. However, iDDC approaches involve long
and complex tasks that often require custom-fit simulators, some coding
expertise, and extensive computing resources. Here we introduce several
resources that offer improved speed and generality, as well as expand
the feasible parameter space for conducting iDDC analyses compared to
other software applications. Specifically, QUETZAL-EGGS are C++ iDDC
simulators; QUETZAL-CRUMBS is a complementary set of Python tools for
simulating on specific landscapes and conducting Approximate Bayesian
Computation (ABC) analyses (e.g., prior sampling, geospatial operations,
ENM/SDM, visualization); DECRYPT is a framework for automated,
biology-informed robustness analysis of the multispecies coalescent
model. All these tools and their dependencies for local use or remote
computations are made readily available in a Docker container package
called QUETZAL-NEST.