geneNetBP-package: Belief Propagation in Genotype-Phenotype Networks

Description Details Author(s) References

Description

The geneNetBP package implements algorithms to infer Conditional Gaussian Bayesian Networks (CG-BN) from genotype-phenotype or Quantitative Trait Loci (QTL) data, absorb evidence in the network and quantify and visualize the changes in network beliefs. Inference of Discrete Bayesian networks is also supported.

Details

Package: geneNetBP
Type: Package
Version: 2.0.4
Date: 2017-02-12
License: GPL-2

This package implements structure learning and conditional probability learning methods from the packages RHugin, gRain and bnlearn to Quantative Trait Loci (QTL) data and provides functions to query and visualize the network beliefs. The function fit.gnbp can be used to learn conditional gaussian bayesian networks from genotype-phenotype data while and fit.dbn allows infering discrete bayesian networks from categorical data. The main function for absorbing evidence and quantifying the changes in beliefs is absorb.gnbp for CG-BN and absorb.dbn for discrete networks. A generic plot method is available for visualizing network beliefs.

For belief propagation in CG-BN, the package makes extensive use of RHugin package that provides an R interface for the Hugin Decision Engine (commercial software for belief propagation). geneNetBP is compatible with demo versino of HuginLite. Note that the demo version of Hugin is restricted to 50 states and 500 cases.The package RHugin is not available on CRAN and can be obtained from R-Forge by accessing the link given below. Hugin Decision Engine should also be installed. Please visit the project homepage for installation instructions.

For discrete bayesian networks, algorithms from the package bnlearn are implemented for structural learning while belief propagation is implemented using gRain.

All the functions in the geneNetBP package are documented. For a complete list of functions, use the command ls(pos = "package:geneNetBP"). A vignette illustrating the use of geneNetBP is available.

Author(s)

Janhavi Moharil <janhavim@buffalo.edu>

Maintainer: Janhavi Moharil <janhavim@buffalo.edu>

References

Janhavi Moharil, Paul May, Daniel P. Gaile, Rachael Hageman Blair (2016). "Belief Propagation in Genotype-Phenotype Networks", Stat Appl Genet Mol Biol, 15(1):39-53. pdf

The geneNetBP Homepage: http://genenetbp.r-forge.r-project.org/

The RHugin Homepage: http://rhugin.r-forge.r-project.org

HUGIN EXPERT website: http://www.hugin.com

HUGIN API Reference Manual


geneNetBP documentation built on May 2, 2019, 5:19 p.m.