Learn R Programming

EDISON (version 1.1.1)

Network Reconstruction and Changepoint Detection

Description

Package EDISON (Estimation of Directed Interactions from Sequences Of Non-homogeneous gene expression) runs an MCMC simulation to reconstruct networks from time series data, using a non-homogeneous, time-varying dynamic Bayesian network. Networks segments and changepoints are inferred concurrently, and information sharing priors provide a reduction of the inference uncertainty.

Copy Link

Version

Install

install.packages('EDISON')

Version

1.1.1

License

GPL-2

Maintainer

Frank Dondelinger

Last Published

March 30th, 2016

Functions in EDISON (1.1.1)

Calculates the MH ratio of the binomial prior.

generateNetwork

Generate a random network.

CalculateChanges

Function to calculate the number of differences between adjaccent network segments.

NetworkRatioBino

Calculates the ratio of binomial prior probabilites.

fix_eigenvalues

Modify network to ensure stationarity.

Initialise the MCMC simulation.

calculateEdgeProbabilitiesSegs

Calculate edge probabilities for fixed segments.

Makes a changepoint shift move.

Sample delta squared.

Calculates the prior probability of the network using the exponential prior.

addProposalNetworkInfo

Add the proposed new network to the new.nets list.

Tune the proposal width for betas.

bp.computeAlpha

Computes the acceptance ratio of two changepoint configurations.

Sample regression coefficients.

Builds response Y and predictor X.

Sample initial sigma squared.

make_structure_move

Makes a structure move.

Main function of the MCMC simulation.

calculateEdgeProbabilities

Calculate the edge probabilities.

Sample initial regression coefficients.

CalculatePriorRatio

Calculates the network prior ratio.

CalculateLikelihoodRatio

Calculates the ratio of two likelihoods in a structure move.

calculateEdgeProbabilitiesTimePoints

Calculate the edge posterior probabilities for each timepoint.

Convert internal representation of networks.

Check if move is acceptable.

Calculates the potential scale reduction factor.

Sample initial parameters for the MCMC simulation.

Makes an exponential hyperparameter move.

Make changepoint birth move.

NetworkProbBino

Calculates the prior probability of the network segments under the binomial prior.

Set the default options for the MCMC simulation.

Wrapper function for starting an MCMC simulation

Compute projection matrix.

Setup and run the MCMC simulation.

Allows for network reconstruction and changepoint detection.

Sample initial number of changepoints.

Calculate inverse gamma distribution.

NetworkRatioExp

Calculates the ratio of exponential network prior probabilities.

psrf_check_hyper

Checks the potential scale reduction factor for the hyperparameters.

Make a network structure or hyperparameter move.

Make changepoint death move.

HyperparameterMove

Make a hyperparameter move.

Collects and saves output.

proposeContinuous

Propose a new real hyperparameter value.

Update sigma squared variances.

Makes a binomial hyperparameter move.

ExpHyperRatioTarget

Calculates the ratio of an exponential hyperparameter move.

CollectNetworkInfo

Collects all the network information in one list.

PriorRatioPoisson

Calculate network prior ratio with Poisson prior.

Sets up initial values of hyperparameters.

simulateNetwork

Generate network and simulate data.

ProposeDiscrete

Propose a new discrete value.

Sample new values for sigma squared.

Check the potential scale reduction factors for all parameters (edges).

calculateCPProbabilities

Calculate the changepoint probabilities.

Samples from the inverse gamma distribution.

calculateCPPGlobal

Calculated the global changepoint probabilities.

Calculate proposal frequencies for changepoint moves.

Read target data.