[R-sig-ME] Performing a single animal model that estimates additive genetic variation for many populations?
Jackie Wood
jackiewood7 at gmail.com
Wed Jun 17 18:47:02 CEST 2015
Hello R-users,
I'm working on a manuscript which investigates the relationship of various
metrics of adaptive potential (additive genetic variation (VA),
heritability (h2), and mean-scaled evolvability (IA)) with population size
in a common garden experiment using a large number of wild, isolated
populations of a vertebrate fish. One comment/request we received on the
manuscript was whether we could fit a single animal model that includes all
of our populations. That is, in addition to the standard terms to estimate
VA, whether we could also include population size and interactions with
population size such that we could examine whether VA (or h2 or IA) varies
among populations, and the magnitude of the interaction between population
and additive (and potentially maternal) effects.
As it currently stands we performed a separate model estimating VA for each
trait for each population which I feel is fairly standard practice, then
simply looked at the correlation between population size and VA (or h2, or
IA) for each trait. If this were a standard mixed-model it would be fairly
easy to incorporate the reviewer's suggestion but of course we are
implementing animal models to estimate our putative metrics of adaptability
and as such we require the use of a pedigree. I'm not sure if it is correct
or possible to construct a single pedigree that incorporates relatedness
information for a large number of populations such that you can estimate VA
for each population and also look at the relationship between VA and
population size in the same model. Is it possible to construct a giant
pedigree where every animal, dam, and sire across all populations get a
unique number but there is an additional column in the pedigree for
population size? Or is there a way to incorporate a large number of
separate pedigrees into a single model that estimates VA for all?
Any advice or insight regarding how to proceed would be greatly appreciated,
Cheers,
Jackie
--
Jacquelyn L.A. Wood, PhD.
Biology Department
Concordia University
7141 Sherbrooke St. West
Montreal, QC
H4B 1R6
Phone: (514) 293-7255
[[alternative HTML version deleted]]
More information about the R-sig-mixed-models
mailing list