compute_risk_score: Get a gene expression matrix with genes as colnames, and...

In christianbromley/raptor: Estimating Renal Anti-Pd(l)1 probabiliTy Of Response scores from gene expression data

Description

This function will score the likelihood of patient response to immune checkpoint blockade based on 4 different pipelines. 1. RAPTOR - based on a set of immune genes 2. COXIS - based on a 25-gene set defined by Bonavtia et al., in review 3. TIS - based on an 18 gene signature from Ayers et al., JCI, 2017 4. Random - a random set of genes to which identical steps were applied as for the other pipelines

Usage

compute_risk_score(expr, coefs, predictor_name, constant)

Arguments

expr	a data frame of gene expression data with genes as column names and samples as rownames. Ideally expression will be in log2(TPM+1) format. The data frame should also include columns for the MSKCC prognostic score if the data is from renal cancer patients (column name = "mskcc").
coefs	a named numeric vector of gene weights.
predictor_name	the name of the probability of response score estimator, for example "RAPTOR"
constant	a constant to add to the scores. We will use the minimum value of the sum of the weighted gene expressions in the training data multiplied by -1 if less than 0.

Details

If more than 50

Value

data.frame with sample ID and column as predictor score

christianbromley/raptor documentation built on Sept. 1, 2020, 12:05 p.m.