Software

nethet: A bioconductor package for high-dimensional exploration of biological network heterogeneity. R package version 1.3.0. Staedler N and Dondelinger F (2015).

NetHMM: Penalized estimation in high-dimensional hidden markov models with state-specific graphical models. The NetHMM-package is a powerful tool for segmenting high-dimensional data. It is based on a hidden markov model with Gaussian emission distributions. Each state is characterized by a mean vector an a covariance matrix. The inverse covariance matrices are sparse and represent a Gaussian graphical models (GGM). Estimation is performed using a novel penalized EM algorithm. Several strategies to select an optimal number of hidden states are implemented. For technical details we refer the reader to: Städler, N. and Mukherjee, S. (2013). Penalized estimation in high-dimensional hidden Markov models with state-specific graphical models. Annals of Applied Statistics 2013, Vol. 7, No. 4, 2157-2179. Manual.

GGMGSA: Gene-set analysis or GSA is a very popular analysis in bioinformatics. Currently available GSA approaches are based on univariate two-sample comparison of single genes. This means that they cannot test for differences in covariance structure between the two conditions. Yet interplay between genes is a central aspect of biological investigation and it is likely that such interplay may differ between conditions. The R package GGMGSA implements a novel approach for gene-set analysis that allows for truly multivariate hypotheses, in particular differences in gene-gene networks between conditions. For technical details of the approach we refer the reader to: Städler, N. and Mukherjee, S. (2013). Network-based multivariate gene-set testing. Preprint arXiv:1308.2771. Manual.

MixGLasso: The R-package MixGLasso is a powerful tool for multivariate, network-based clustering of high-dimensional data. This approach simultaneously identifies clusters and learns networks topologies. MixGLasso is based on a Gaussian mixture models. Estimation is performed using a EM-type algorithm which uses L1-penalization to regularize estimation and takes care of scaling and related issues that arise due to the unknown nature of the cluster assignments. For technical details we refer the reader to: Städler, N. and Mukherjee, S. (2013). Penalized estimation in high-dimensional hidden Markov models with state-specific graphical models. Annals of Applied Statistics 2013, Vol. 7, No. 4, 2157-2179. Manual.

DiffNet: Network inference is subject to statistical uncertainty and observed differences between two networks inferred from two datasets may be due to noise and variability in estimation rather than any true difference in underlying network topology. Significance testing for network differences is a challenging statistical problem, involving high-dimensional estimation and comparison of non-nested hypotheses. Our recently developed method “differential network” performs formal two-sample testing between high-dimensional Gaussian graphical models (GGMs) and is implemented in the R-package DiffNet. For technical details of the approach we refer the reader to: Städler, N. and Mukherjee, S. (2013). Two-Sample Testing in High-Dimensional Models. Preprint arXiv:1210.4584. Manual.

DiffRegr: The R-package DiffRegr performs formal two-sample testing between high-dimensional regression models. For technical details of the approach we refer the reader to: Städler, N. and Mukherjee, S. (2013). Two-Sample Testing in High-Dimensional Models. Preprint arXiv:1210.4584. Manual.

fmrlasso: An R-package for the computation of the FMRLasso (Discussion paper in TEST, 2010, Volume 19, 209-285) estimator. For more information and examples read the Manual.

Advertisements