software

ActiveDriver looks for significant enrichment of cancer mutations in signalling sites of proteins to discover driver genes and mechanisms
ActiveDriver looks for significant enrichment of cancer mutations in signalling sites of proteins to discover driver genes and mechanisms

ActiveDriver – ActiveDriver is a mutation analysis tool that discovers cancer driver genes with frequent mutations in protein signalling sites such as post-translational modifications (phosphorylation, ubiquitination, etc). The model identifies genes where cancer mutations in signalling sites are more frequent than expected from the sequence of the entire gene. Integration of mutations with signalling information helps find new driver genes and propose candidate mechanisms to known drivers. ActiveDriver is a generalised linear regression model that is available as an R package. The software package with a tutorial and relevant datasets can be found here.
Publication: Reimand J, Bader GD. Systematic analysis of somatic mutations in phosphorylation signaling predicts novel cancer drivers. Mol Syst Biol. 2013;9:637. doi: 10.1038/msb.2012.68. PubMed PMID: 23340843; PubMed Central PMCID: PMC3564258.

MIMP software uses machine learning to predict the impact of cancer SNVs on kinase binding sites
MIMP software uses machine learning to predict the impact of cancer SNVs on kinase binding sites

MIMP – MIMP is a mutation analysis tool that identifies mutations that interfere with small sites in protein sequences. MIMP uses machine learning to detect sequence motifs involved in kinase binding and phosphorylation. It then predicts whether cancer mutations in phosphorylation sites remove existing motifs or create new motifs, potentially causing rewiring of kinase signalling networks. MIMP uses Gaussian mixture models and Bayesian statistics to compute posterior probabilities of mutation impact on kinase binding sites. The development of MIMP was led by Omar Wagih as his Master’s project. MIMP is available as an interactive web site and an R package at http://mimp.baderlab.org.
Publication (PDF): Wagih O, Reimand J*, Bader GD*. MIMP: predicting the impact of mutations on kinase-substrate phosphorylation. Nat Methods. 2015 Jun;12(6):531-3. doi:10.1038/nmeth.3396. Epub 2015 May 4. PubMed PMID: 25938373.

HyperModules analyses protein interaction networks to discover subnetworks where gene mutations are correlated with clinical characteristics and patient survival.
HyperModules analyses protein interaction networks to discover subnetworks where gene mutations are correlated with clinical characteristics and patient survival.

HyperModules – HyperModules is a tool for analysing biological networks with disease mutations and clinical information. HyperModules helps discover gene modules (i.e. subnetworks) where disease mutations correlate with patient survival or other clinical characteristics. This is achieved with a greedy local network search strategy that identifies survival correlations in subnetworks spanning from individual genes in the network. A network-based permutation test is then used to compute significance of detected modules and filter false positive results. HyperModules is available as a Cytoscape app whose development was led my Alvin Leung as a Google Summer of Code project. The Hypermodules app can be downloaded from Cytoscape App Store and a tutorial is also available.
Publication: Leung A, Bader GD*, Reimand J*. HyperModules: identifying clinically and phenotypically significant network modules with disease mutations for biomarker discovery. Bioinformatics. 2014 Aug 1;30(15):2230-2. doi: 10.1093/bioinformatics/btu172. Epub 2014 Apr 8. PubMed PMID: 24713437; PubMed
Central PMCID: PMC4103591.

g:Profiler analyses gene lists with current knowledge of gene function and determines the most characteristic biological pathways and processes.
g:Profiler analyses gene lists with current knowledge of gene function and determines the most characteristic biological pathways and processes.

g:Profiler – g:Profiler is a web server for performing pathway enrichment analysis of gene lists. It detects the most characteristic biological pathways, processes, molecular functions and many other features in a gene list of interest using information from Gene Ontology, pathway databases and other resources. g:Profiler applies Fisher’s exact tests to determine these significant characteristics, and introduces a custom multiple testing procedure to better filter false positive results. It also includes a method to analyse ranked gene lists that usually provides a more robust pathway analysis with higher level of detail. g:Profiler supports analysis of more than 200 species and provides several other useful tools. Notably the g:Convert service helps translate between hundreds of gene identifiers, symbols, and accession numbers. g:Profiler is available as an interactive website at http://biit.cs.ut.ee/gprofiler and as an R package available here.
Publication: Reimand J, Kull M, Peterson H, Hansen J, Vilo J. g:Profiler–a web-based toolset for functional profiling of gene lists from large-scale experiments. Nucleic Acids Res. 2007 Jul;35(Web Server issue):W193-200. Epub 2007 May 3. PubMed PMID: 17478515; PubMed Central PMCID: PMC1933153.