Tutorial on survival modelling with omics data
Identification of genomic, molecular and clinical markers predictive of patient survival is important for developing personalized disease prevention, diagnostic and treatment approaches. Modern omics technologies have made it possible to investigate the prognostic impact of markers at multiple molecular levels, including genotype, DNA methylation, transcriptomics, proteomics and metabolomics, and how these risk factors complement clinical characterization of patients for prognostic prediction. However, the massive omics data pose challenges for studying relationships between the molecular information and patients' survival outcomes. We demonstrate a general workflow of survival analysis, with emphasis on dealing with high-dimensional omics features, using both univariate and multivariate approaches. In particular, we describe commonly used Cox-type penalized regressions and Bayesian models for feature selection in survival analysis with multi-omics data, where caution is needed to account for the underlying structure both within and between omics data sets. A step-by-step R tutorial using TCGA survival and omics data for the execution and evaluation of survival models has been made available at https://ocbe-uio.github.io/survomics/survomics.html.
READ FULL TEXT