runSegratioMM function

Run a Bayesian mixture model for marker dosage with minimal effort

Given segregation ratios and a ploidy level, a mixture model is constructed with default priors and initial values and JAGS run to produce an MCMC sample for statistical inference. Returns an object of S3 class runJagsWrapper

runSegratioMM(seg.ratios, model, priors = setPriors(model),
 inits = setInits(model, priors), jags.control =
 setControl(model, stem, burn.in = burn.in, sample = sample, thin = thin),
 burn.in = 2000, sample = 5000, thin = 1, stem = "test", fix.one = TRUE,
 print = TRUE, plots = TRUE, print.diagnostics = TRUE,
 plot.diagnostics = TRUE, run.diagnostics.later=FALSE )

Arguments

• seg.ratios: Object of class segRatio

  contains the segregation ratios for dominant markers and other information such as the number of dominant markers per individual

• model: object of class modelSegratioMM specifying model parameters, ploidy etc

• priors: object of class priorsSegratioMM indicating priors that are vague , strong or specified

• inits: A list of initial values usually produced by setInits

• jags.control: Object of class jagsControl containing MCMC burn in, sample and thinning as well as relavant files for BUGS commands, inits and data

• burn.in: size of MCMC burn in (Default: 2000)

• sample: size of MCMC sample (default: 5000)

• thin: thinning interval between consecutive observations (default: 1 or no thinning)

• stem: text to be used as part of JAGS .cmd file name

• fix.one: Logical to fix the dosage of the observation closest to the centre of each component on the logit scale. This can greatly assist with convergence (Default: TRUE)

• print: logical for printing monitoring and summary information (default: TRUE)

• plots: logical to plotting MCMC posterior distributions (default: TRUE)

• print.diagnostics: logical for printing disagnostic statistics (default: TRUE)

• plot.diagnostics: logical for diagnostic plots (default: TRUE)

• run.diagnostics.later: should diagnostics be run later which may help if there are convergence problems (Default: FALSE)

Returns

Returns object of class runJagsWrapper with components - seg.ratios: Object of class segRatio

contains the segregation ratios for dominant markers

• model: object of class modelSegratioMM specifying model parameters, ploidy etc

• priors: Object of class priorsSegratioMM specifying prior distributions

• inits: A list of initial values usually produced by setInits

• jags.control: Object of class jagsControl containing MCMC burn in, sample and thinning as well as relavant files for BUGS commands, inits and data

• stem: text to be used as part of JAGS .cmd file name and other files

• fix.one: Logical to fix the dosage of the observation closest to the centre of each component on the logit scale. This can greatly assist with convergence (Default: TRUE)

• run.jags: object of class runJAGS produced by running JAGS

• mcmc.mixture: Object of type segratioMCMC

  produced by coda usually by using readJags

• diagnostics: list containing various diagnostic summaries and statistics produced by coda

• summary: summaries of posterior distributions of model parameters

• doses: object of class dosagesMCMC containing posterior probabilities of dosages for each marker dosage and allocated dosages

• DIC: Deviance Information Critereon

Author(s)

Peter Baker p.baker1@uq.edu.au

See Also

setPriors setInits

expected.segRatio

segRatio

setControl

dumpData dumpInits and diagnosticsJagsMix

Examples

## simulate small autooctaploid data set
a1 <- sim.autoMarkers(8,c(0.7,0.2,0.1),n.markers=100,n.individuals=50)
##print(a1)
sr <-  segregationRatios(a1$markers)
x <- setModel(3,8)

## Not run:

## fit simple model in one hit

x.run <- runSegratioMM(sr, x, burn.in=200, sample=500)
print(x.run)
## End(Not run)