How to find coefficients of all genes for prediction of risk score with some threshold using multivariate cox regression model in TCGA-LIHC dataset ?

Question

[donnieDarko]

Hello,

I have seen many articles which have analyzed TCGA datasets using univariate/ multivariate cox regression model expressing risk score as a linear formula of coef1 * exp.gene1 + coef2 * exp.gene2 + coef3 * exp.gene3 + ...... upto some p.value threshold. I want to find those coefficients for genes, in TCGA-LIHC dataset and develop the risk score model, which are significantly correlated with the patient's overall survival.

How do I find the coefficients and the risk score formula ?

[Nobody]

Hi,

To find coefficients, you should do multivariate cox analysis. but first, you have to perform a univariate test to choose those that have a prognostic effect