Bioconductor-mirror/genphen: A tool for quantification of associations between genotypes and phenotypes with statistical learning techniques such as random forests and support vector machines as well as with Bayesian inference using hierarchical models

Genetic association studies have become an essential tool for studying the relationship between genotypes and phenotypes. They are necessary for the discovery of disease-causing genetic variants. Here we provide a tool for conducting genetic association studies, which uses statistical learning techniques such as random forests and support vector machines, as well as using Bayesian inference with Bayesian hierarchical models. These techniques are superior to the commonly used (frequentist) statistical approaches, alleviating the multiple hypothesis problems and the need for P value corrections, which often lead to massive numbers of false negatives. Thus, with genphen we provide a framework to compare the results obtained using frequentist methods with those obtained using the more sophisticated methods provided by this tool. The tool also provides a few visualization functions which enable the user to inspect the results of such genetic association study and conveniently select the genotypes which have the highest strength of association with the phenotype.

Getting started

Package details

AuthorSimo Kitanovski
Bioconductor views Bayesian Classification FeatureExtraction Genetics GenomeWideAssociation Regression SequenceMatching Sequencing SupportVectorMachine
MaintainerSimo Kitanovski <[email protected]>
LicenseGPL (>= 2)
Package repositoryView on GitHub
Installation Install the latest version of this package by entering the following in R:
Bioconductor-mirror/genphen documentation built on June 1, 2017, 9:43 a.m.