RLassoCox

The ASURI (Analysis of SUrvival and patients RIsk prediction based on gene signatures) package discovers marker genes that are related to risk prediction capabilities and to a clinical variable of interest. It uses two main steps, including subsampling glmnet and unicox. The package implements robust functions to discover survival markers related to a clinical phenotype and to predict a risk score, allowing to study the patient's risk based on the gene signatures. Several plots are provided to visualise the relevance of the genes, the risk score, and patient stratification, as well as a robust version of the Kaplan-Meier curves.

pathwayPCA is an integrative analysis tool that implements the principal component analysis (PCA) based pathway analysis approaches described in Chen et al. (2008), Chen et al. (2010), and Chen (2011). pathwayPCA allows users to: (1) Test pathway association with binary, continuous, or survival phenotypes. (2) Extract relevant genes in the pathways using the SuperPCA and AES-PCA approaches. (3) Compute principal components (PCs) based on the selected genes. These estimated latent variables represent pathway activities for individual subjects, which can then be used to perform integrative pathway analysis, such as multi-omics analysis. (4) Extract relevant genes that drive pathway significance as well as data corresponding to these relevant genes for additional in-depth analysis. (5) Perform analyses with enhanced computational efficiency with parallel computing and enhanced data safety with S4-class data objects. (6) Analyze studies with complex experimental designs, with multiple covariates, and with interaction effects, e.g., testing whether pathway association with clinical phenotype is different between male and female subjects. Citations: Chen et al. (2008) <https://doi.org/10.1093/bioinformatics/btn458>; Chen et al. (2010) <https://doi.org/10.1002/gepi.20532>; and Chen (2011) <https://doi.org/10.2202/1544-6115.1697>.

The iNETgrate package provides functions to build a correlation network in which nodes are genes. DNA methylation and gene expression data are integrated to define the connections between genes. This network is used to identify modules (clusters) of genes. The biological information in each of the resulting modules is represented by an eigengene. These biological signatures can be used as features e.g., for classification of patients into risk categories. The resulting biological signatures are very robust and give a holistic view of the underlying molecular changes.

Motivation: The understanding of cancer mechanism requires the identification of genes playing a role in the development of the pathology and the characterization of their role (notably oncogenes and tumor suppressors). Results: We present an R/bioconductor package called MoonlightR which returns a list of candidate driver genes for specific cancer types on the basis of TCGA expression data. The method first infers gene regulatory networks and then carries out a functional enrichment analysis (FEA) (implementing an upstream regulator analysis, URA) to score the importance of well-known biological processes with respect to the studied cancer type. Eventually, by means of random forests, MoonlightR predicts two specific roles for the candidate driver genes: i) tumor suppressor genes (TSGs) and ii) oncogenes (OCGs). As a consequence, this methodology does not only identify genes playing a dual role (e.g. TSG in one cancer type and OCG in another) but also helps in elucidating the biological processes underlying their specific roles. In particular, MoonlightR can be used to discover OCGs and TSGs in the same cancer type. This may help in answering the question whether some genes change role between early stages (I, II) and late stages (III, IV) in breast cancer. In the future, this analysis could be useful to determine the causes of different resistances to chemotherapeutic treatments.

glmSparseNet is an R-package that generalizes sparse regression models when the features (e.g. genes) have a graph structure (e.g. protein-protein interactions), by including network-based regularizers. glmSparseNet uses the glmnet R-package, by including centrality measures of the network as penalty weights in the regularization. The current version implements regularization based on node degree, i.e. the strength and/or number of its associated edges, either by promoting hubs in the solution or orphan genes in the solution. All the glmnet distribution families are supported, namely "gaussian", "poisson", "binomial", "multinomial", "cox", and "mgaussian".

Reconstructing gene regulatory networks and transcription factor activity is crucial to understand biological processes and holds potential for developing personalized treatment. Yet, it is still an open problem as state-of-art algorithm are often not able to handle large amounts of data. Furthermore, many of the present methods predict numerous false positives and are unable to integrate other sources of information such as previously known interactions. Here we introduce KBoost, an algorithm that uses kernel PCA regression, boosting and Bayesian model averaging for fast and accurate reconstruction of gene regulatory networks. KBoost can also use a prior network built on previously known transcription factor targets. We have benchmarked KBoost using three different datasets against other high performing algorithms. The results show that our method compares favourably to other methods across datasets.