SSCox is an R package to perform a survival analysis of patients based on their genomic profiles using sparse-smoothed Cox proportional hazard model. The model is described in the manuscript Sparse-smoothed Cox proportional hazard model for genomic profiles. The R package has currently been tested only in Linux/Unix environment. Download the R package here (version 1.0, Jan 2017). The description of the R package can be found here.
The statistical model is suitable for genomic data (or any other data in general) where the predictors are relatively inter-dependent. We describe below the data preparation for copy number alteration (CNA) profile data that we use in the manuscript. In general, the data can be organised such that the rows are the patients/individuals and the columns are the predictors or genomic locations.
The input data that we use in the manuscript are basically the output of CNAnorm across different sequences/patients. After you normalise the CNA (per patient) using the CNAnorm package, save the results, preferably with .norm file extension.
R CMD INSTALL SSCox_[version].tar.gz
You can also install SSCox in your own directory by using
R CMD INSTALL -l your-own-directory SSCox_[version].tar.gz
if you have installed the package in your own directory (not in the default location).
Last modified: Wed Jan 4 18:52:05 GMT 2017