Leo Lahti*, Olli-Pekka Huovilainen, António Gusmão and Juuso Parkkinen (*) University of Helsinki, Finland; corresponding author NetResponse provides probabilistic tools for functional analysis of interaction networks, and provides a global map of network activation patterns across versatile contexts. The implementations are based on probabilistic models and variational learning to deal rigorously with inherent uncertainty in real-life data, and to allow incorporation of prior information in the analysis. Currently available tools: NetResponse¹ to detect and characterize local, context-specific activation patterns and to construct global functional maps of large interaction networks by utilizing functional information of the network based on non-parametric analysis with variational Dirichlet process Gaussian mixture models². Interaction Component Model (ICMg³): to discover functional network modules, or communities, taking into account the uncertainty in network structure. Module discovery can be supervised by functional information of the network. Visualization: Preliminary tools are available to visualize netresponse output. Install the netresponse visualization package and check the package vignette (in package source, inst/doc/) for examples. Applicability of the models has been demonstrated by case studies in computational biology (Lahti et al., 2010, Parkkinen and Kaski, 2010), where the algorithms have been used to investigate the structure and context-specific transcriptional activity of genome-scale interaction networks in human body. The techniques are implemented in R, an open source environment for statistical computing. For installation instructions, see the project page at BioConductor. For further documentation, see the package vignette. In addition, Matlab source code for the NetResponse algorithm¹ is available.

Licensing terms

Licensed under an open source software license (GNU GPL >=2). Your feedback and contributions are welcome.

References

1. Leo Lahti et al. (2010). Global modeling of transcriptional responses in interaction networks. Bioinformatics. (abstract; advanced online-access (pdf); preprint with supplementary figures (pdf)

2. Kenichi Kurihara et al. (2007). Accelerated variational Dirichlet process mixtures. In B. Schölkopf, J. Platt, and T. Hoffman, eds., Advances in Neural Information Processing Systems 19, 761-768. MIT Press, Cambridge, MA.
3. Juuso Parkkinen and Samuel Kaski. Searching for functional gene modules with interaction component models. BMC Systems Biology 2010, 4:4.