dnet: Integrative analysis of omics data in terms of network, evolution and ontology

This package is initiated to fill in the need of an open-source tool for high-throughput omics data in an integrative manner in terms of network, evolution and ontology. More specifically, dnet intends to analyse the biological network whose nodes/genes are associated with digitised information such as expression levels across samples. To help make sense of identified networks, enrichment analysis is also supported using a wide variety of pre-compiled ontologies and phylostratific gene age information in major organisms including human, mouse, rat, chicken, C.elegans, fruit fly, zebrafish and arabidopsis. Add-on functionalities are supports for semantic similarity calculation between ontology terms (and genes), and network affinity calculation using Random Walk with Restart; both can be done via high-performance parallel computing. In summary, dnet aims to deliver an eye-intuitive tool with rich visuals but less inputs.

Version: 1.0.6
Depends: R (≥ 3.1.0), igraph, supraHex
Imports: graph, Rgraphviz, Matrix, Biobase
Suggests: limma, survival, foreach, doMC
OS_type: unix
Published: 2014-09-22
Author: Hai Fang and Julian Gough
Maintainer: Hai Fang <hfang at cs.bris.ac.uk>
License: GPL-2
URL: http://supfam.org/dnet
NeedsCompilation: no
Citation: dnet citation info
Materials: NEWS
CRAN checks: dnet results


Reference manual: dnet.pdf
Package source: dnet_1.0.6.tar.gz
Windows binaries: r-devel: dnet_1.0.6.zip, r-release: dnet_1.0.6.zip, r-oldrel: dnet_1.0.0.zip
OS X Snow Leopard binaries: r-release: dnet_1.0.6.tgz, r-oldrel: not available
OS X Mavericks binaries: r-release: dnet_1.0.6.tgz
Old sources: dnet archive

Reverse dependencies:

Reverse depends: dcGOR