ProgPredLasso {PPLasso}

Identification of prognostic and predictive biomarkers

Description

The computes the regularization path of the Prognostic Predictive Lasso described in the paper Zhu et al. (2022) given in the references.

Usage

ProgPredLasso(X1, X2, Y=Y, cor_matrix=NULL, gamma=0.99, maxsteps=500, lambda='single')

Arguments

X1

Design matrix of patients characteristics with treatment 1

X2

Design matrix of patients characteristics with treatment 2

Y

Response variable

cor_matrix

Correlation matrix of biomarkers. If not specified, the function cvCovEst from package cvCovEst will be used to estimate this matrix.

gamma

Parameter \gamma defined in the paper Zhu et al. (2020) given in the references. Its default value is 0.99.

maxsteps

Integer specifying the maximum number of steps for the generalized Lasso algorithm. Its default value is 500.

lambda

Using single tuning parameter or both.

Value

Returns a list with the following components

lambda

different values of the parameter \lambda considered.

beta

matrix of the estimations of \beta for all the \lambda considered.

beta.min

estimation of \beta which minimize the MSE.

bic

BIC for all the \lambda considered.

mse

MSE for all the \lambda considered.

Author(s)

Wencan Zhu, Celine Levy-Leduc, Nils Ternes