startPed {MasterBayes} | R Documentation |
An object containing the starting parameterisation of a model, and logical variables indicating wether parameters should be estimated or fixed at the starting parameterisation. By default the starting parameterisation is obtained through a mixture of Maximum Likelihood and heuristic techniques.
startPed(id=NULL, G=NULL, estG=TRUE, A=NULL, estA=TRUE, E1=NULL, estE1=TRUE, E2=NULL, estE2=TRUE, dam=NULL,sire=NULL, estP=TRUE, beta=NULL, estbeta=TRUE, USdam=NULL, estUSdam=TRUE, USsire=NULL, estUSsire=TRUE, ...)
id |
vector of indivual id's for G , dam and sire |
G |
list of genotype objects |
estG |
logical; should genotypes be estimated? |
A |
list of allele frequencies |
estA |
logical; should base-population allele frequencies be estimated? |
E1 |
vector of allelic dropout rates for codominat markers or probability of mis-scoring a dominant allele as recessive for dominant markers (default=0.005) |
estE1 |
logical; should allelic dropout rates be estimated? |
E2 |
vector of stochastic genotyping error rates for codominat markers or probability of mis-scoring a recessive allele as dominant for dominant markers (default=0.005) |
estE2 |
logical; should stochastic error rates be estimated? |
dam |
vector of dam's. If the dam is unknown use NA . All dam's must be in id |
sire |
vector of sire's. If the sire is unknown use NA . All sire's must be in id |
estP |
logical; should the pedigree be estimated? |
beta |
vector of population-level parameters |
estbeta |
logical; should the population-level parameters be estimated? |
USdam |
vector of unsampled female population sizes |
estUSdam |
logical; should the female population sizes be estimated? |
USsire |
vector of unsampled male population sizes |
estUSsire |
logical or character; if TRUE the male population size is estimated separately from the female population size, if "USdam" male and female population sizes are constrained to be the same. |
... |
If estG=FALSE
an approximation is used for genotyping error. In this case error rates and allele frequencies are not estimated but fixed at the starting parameterisation. If indivdiuals have been typed more than once, then the approxiamtion only uses the genotype that first appears in the GdP$G
object passed to MCMCped
. If A
is not specified estimates are taken directly from GdP$G
using extractA
. If E1
and E2
are not specified they are set to 0.005. Note that if the approximation for genotyping error is used with codominant markers, Wang's (2005) model is not used, and the CEVUS model (Marshall 1998) is adopted. In this case E2
is the per-allele error rate and E2
(2-E2
) is the per-genotype error rate used by CERVUS. If dam
and sire
are not specified the most likely set of parents given the genetic data are used (see MLE.ped
). The starting value of beta
, if not given, is the MLE of beta given the starting pedigree (see MLE.beta
). The starting values of USdam
and USsire
, if not given, are the MLE based on the genotype data (see MLE.popsize
).
list containing the arguments passed
Jarrod Hadfield j.hadfield@ed.ac.uk
# In this example we simulate a pedigree and then fix the # pedigree and estimate the population level paarmeters data(WarblerP) var1<-expression(varPed(c("lat", "long"), gender="Male", relational="OFFSPRING")) # paternity is to be modelled as a function of distance # between offspring and male territories res1<-expression(varPed("offspring", restrict=0)) # indivdiuals from the offspring generation are excluded as parents res2<-expression(varPed("terr", gender="Female", relational="OFFSPRING", restrict="==")) # mothers not from the offspring territory are excluded PdP<-PdataPed(formula=list(var1,res1,res2), data=WarblerP, USsire=FALSE) simped<-simpedigree(PdP, beta=-0.25) # simulate a pedigree where paternity drops with distance (beta=-0.25) sP<-startPed(dam=simped$ped[,2], sire=simped$ped[,3], estP=FALSE) model1<-MCMCped(PdP=PdP, sP=sP, nitt=3000, thin=2, burnin=1000) plot(model1$beta) # The true underlying value is -0.25