Package 'sGBJ'

Title: Survival Extension of the Generalized Berk-Jones Test
Description: Implements an extension of the Generalized Berk-Jones (GBJ) statistic for survival data, sGBJ. It computes the sGBJ statistic and its p-value for testing the association between a gene set and a time-to-event outcome with possible adjustment on additional covariates. Detailed method is available at Villain L, Ferte T, Thiebaut R and Hejblum BP (2021) <doi:10.1101/2021.09.07.459329>.
Authors: Laura Villain [aut, cre], Thomas Ferte [aut], Rodolphe Thiebault [aut], Boris P. Hejblum [aut]
Maintainer: Laura Villain <[email protected]>
License: GPL (>= 3)
Version: 0.1.0
Built: 2025-02-02 05:41:36 UTC
Source: https://github.com/lauravillain/sgbj

Help Index

.epsilon_matrix

Description

Compute the epsilon matrix by permutation for the sGBJ_scores() function.

Usage

.epsilon_matrix(Z, nperm, surv, factor_matrix, covariates = NULL, dat)

Arguments

Z

the score vector returned by .survival_scores() function.
nperm

number of permutations performed to estimate the epsilon matrix. Default is 300.
surv

a Surv object of length n
factor_matrix

a n x p data.frame of the expression for the particular gene set of interest being tested
covariates

a n x l matrix of the covariates to adjust upon. Default is NULL
dat

data used to fit survival model returned by .survival_scores() function.

Value

The epsilon matrix.

.survival_scores

Description

Compute the survival score

Usage

.survival_scores(factor_matrix, covariates = NULL, surv)

Arguments

factor_matrix

a n x p data.frame of the expression for the particular gene set of interest being tested
covariates

a matrix nxl of the covariates to adjust. Default is NULL
surv

a Surv object of length n

Value

A list of length 3 with the updated factor_matrix (same as factor_matrix but removing columns for which survival model failed to converge), the Z matrix and the data used to fit survival model.

A data file used for testing sGBJ

Description

A data file used for testing sGBJ

Compute the sGBJ statistic and its p-value quantifying a gene set expression association with survival

Description

This function is the main function of the sGBJ package to perform Gene Set Analysis in the context of time-to-event outcome.

Usage

sGBJ(surv, factor_matrix, covariates = NULL, nperm = 300)

Arguments

surv

a Surv object of length n
factor_matrix

a n x p data.frame of the expression for the particular gene set of interest being tested
covariates

a n x l matrix of the covariates to adjust upon. Default is NULL
nperm

number of permutations performed to estimate the epsilon matrix. Default is 300.

Value

The sGBJ statistic and its associated p-value associated

Examples

n <- 100
surv_data <- data.frame(Time = runif(n = n, min = 0, max = 100),
                        event = rbinom(n = n, size = 1, prob = 0.5))
surv <- survival::Surv(time = surv_data$Time, event = surv_data$event)

factor_matrix <- data.frame(P1 = rnorm(n = n),
                             P2 = rnorm(n = n))

sGBJ::sGBJ(surv,factor_matrix, nperm = 2)

Compute the sGBJ statistic along with its p-value quantifying the association between a gene set and survival outcome

Description

Compute the sGBJ statistic along with its p-value quantifying the association between a gene set and survival outcome

Usage

sGBJ_scores(surv, factor_matrix, covariates = NULL, nperm = 300)

Arguments

surv

a Surv object of length n
factor_matrix

a n x p data.frame of the expression for the particular gene set of interest being tested
covariates

a n x l matrix of the covariates to adjust upon. Default is NULL
nperm

number of permutations performed to estimate the epsilon matrix. Default is 300.

Value

a list containing the sGBJ statistic estimation and its associated p-value

Examples

n <- 100
 surv_data <- data.frame(Time = runif(n = n, min = 0, max = 100),
                         event = rbinom(n = n, size = 1, prob = 0.5))
 surv <- survival::Surv(time = surv_data$Time, event = surv_data$event)

 factor_matrix <- data.frame(P1 = rnorm(n = n),
                              P2 = rnorm(n = n))

 sGBJ::sGBJ_scores(surv,factor_matrix, nperm = 2)

 # with covariates

 covariates <- data.frame(age = runif(n = n, 60, 90))

 sGBJ_scores(surv,factor_matrix, nperm = 2, covariates = covariates)

surv_calc_scores_stats

Description

An adaptation of GBJ::calc_scores_stats() to survival context. Wrapper of sGBJ_scores() function.

Usage

surv_calc_scores_stats(null_model, factor_matrix, nperm = 300)

Arguments

null_model

An R cox model fitted with survival::coxph().
factor_matrix

An n x p matrix with each factor as one column. There should be no missing data.
nperm

Number of permutations (default is 300)

Value

A list with the elements:

test_stats

The p score test statistics.
cor_mat

The p x p matrix giving the pairwise correlation of every test statistic pairs.

Examples

n <- 100
surv_data <- data.frame(Time = runif(n = n, min = 0, max = 100),
                        event = rbinom(n = n, size = 1, prob = 0.5))
surv <- survival::Surv(time = surv_data$Time, event = surv_data$event)

factor_matrix <- data.frame(P1 = rnorm(n = n),
                             P2 = rnorm(n = n))

covariates <- data.frame(age = runif(n = n, 60, 90))

null_model <- survival::coxph(surv ~ age, data = covariates, x = TRUE)
surv_reg_stats <- surv_calc_scores_stats(null_model = null_model,
                                         factor_matrix = factor_matrix,
                                         nperm = 2)#nperm = 300)

GBJ::GBJ(test_stats=surv_reg_stats$test_stats, cor_mat=surv_reg_stats$cor_mat)