GPrank: Gaussian Process Ranking of Multiple Time Series
Version 0.1.2

Implements a Gaussian process (GP)-based ranking method which can be used to rank multiple time series according to their temporal activity levels. An example is the case when expression levels of all genes are measured over a time course and the main concern is to identify the most active genes, i.e. genes which show significant non-random variation in their expression levels. This is achieved by computing Bayes factors for each time series by comparing the marginal likelihoods under time-dependent and time-independent GP models. Additional variance information from pre-processing of the observations is incorporated into the GP models, which makes the ranking more robust against model overfitting. The package supports exporting the results to 'tigreBrowser' for visualisation, filtering or ranking.

Package details

AuthorHande Topa [aut, cre], Antti Honkela [aut]
Date of publication2016-12-21 17:02:00
MaintainerHande Topa <[email protected]>
LicenseMIT + file LICENSE
Version0.1.2
URL https://github.com/PROBIC/GPrank
Package repositoryView on CRAN
Installation Install the latest version of this package by entering the following in R:
install.packages("GPrank")

Try the GPrank package in your browser

Any scripts or data that you put into this service are public.

GPrank documentation built on May 29, 2017, 5:51 p.m.