GRENITS (version 1.24.0)

mcmc.defaultParams_gauss: Default Parameters for Linear Model with Gaussian distributed replicates

Description

Create parameter vector with default parameters for ReplicatesNet_gauss function

Usage

mcmc.defaultParams_gauss()

Value

Returns a single vector with the following elements (in this order):

(1) samples

Number of MCMC iterations to run

(2) burn.in

Number of initial iterations to discard as burn in

(3) thin

Subsampling frequency

(4) c

Shape parameter 1 for Beta(c,d) prior on rho (connectivity parameter)

(5) d

Shape parameter 2 for Beta(c,d) prior on rho (connectivity parameter)

(6) sigma.s

Standard deviation parameter for N(0,sigma.s) prior on B (Regression coefficients)

(7) a

Shape parameter for Gamma(a,b) prior on lambda (Regression precision)

(8) b

Rate parameter for Gamma(a,b) prior on lambda (Regression precision)

(9) a_exp

Shape parameter for Gamma(a_exp,b_exp) prior on tau (Replicates precision)

(10) b_exp

Rate parameter for Gamma(a_exp,b_exp) prior on tau (Replicates precision)

(11) sigma.mu

Standard deviation parameter for N(0,sigma.mu) prior on mu (Regression intercept)

(12) fix.y.iter

Number of iterations for which sampled data Y is fixed

Details

Use this function to generate a template parameter vector to use non-default parameters for the ReplicatesNet_gauss model.

References

Morrissey, E.R., Juarez, M.A., Denby, K.J. and Burroughs, N.J. 2010. On reverse engineering of gene interaction networks using time course data with repeated measurements. Bioinformatics 2010; doi: 10.1093/bioinformatics/btq421

See Also

plotPriors, ReplicatesNet_gauss.

Examples

    # Get default parameters
    linearNet_Gauss.params <- mcmc.defaultParams_gauss()

    # Change run length
    linearNet_Gauss.params[1] <- 200000

    # Change prior regression precision 
    linearNet_Gauss.params[7] <- 0.001
    linearNet_Gauss.params[8] <- 0.001

    # Plot to visualise changes
    plotPriors(linearNet_Gauss.params)

    ## Use to run ReplicatesNet_gauss ...