SignS: a parallelized, open-source, freely available, web-based tool for gene selection and molecular signatures for survival and censored data.

摘要: Censored data are increasingly common in many microarray studies that attempt to relate gene expression patient survival. Several new methods have been proposed the last two years. Most of these methods, however, not available biomedical researchers, leading re-implementations from scratch ad-hoc, and suboptimal, approaches with survival data. We developed SignS (Signatures for Survival data), an open-source, freely-available, web-based tool R package selection, building molecular signatures, prediction implements four which, according existing reviews, perform well and, by being a very different nature, offer complementary approaches. use parallel computing via MPI, large decreases user waiting time. Cross-validation is used asses predictive performance stability solutions, latter issue increasing concern given there often several solutions similar performance. Biological interpretation results enhanced because genes signatures models can be sent other freely-available on-line tools examination PubMed references, GO terms, KEGG Reactome pathways selected genes. first analysis data, one few researchers as target users. also bioinformatics applications extensively parallelization, including fault tolerance crash recovery. Because its combination implemented, usage computing, code availability, links additional bases, unique tool, will immediate relevance biostatisticians bioinformaticians.