BCMC

This R package implements BCMC (Biomarker Categorization in Meta-analysis by Concordance) for biomarker detection and categorization. The details of this method can be found in our paper.

Installation

To install the BCMC package, you will first need to install devtools package and then execute the following code:

devtools::install_github('kehongjie/BCMC')

Main Functions

There are three main functions in this package:

bcmc runs the BCMC and return the meta analysis statistics and predicted weight patterns for each gene.
perm.bcmc uses a permutation-based test to calculate the p-value and FDR adjusted p-value (also known as q-value) after doing the BCMC.
comp.bcmc is the combination of above two functions and runs the complete procedure of BCMC, which includes the calculation of statistics, the prediction of weight pattern, and the calculation of p-value and q-value.

You can always use the following command to see more details:

library(BCMC)
?bcmc
?perm.bcmc
?comp.bcmc

Data and Examples

This package also includes three data sets under the /data folder:

SimulDE.RData: A simulated DE data with 2000 genes and 5 studies.
PanGyn.RData: A TCGA Pan Gynecologic cancer data with coding genes only.
PanKidney.RData: A TCGA Pan Kidney cancer data that include mRNA, miRNA as well as lncRNA.

Here is a toy example of running the comp.bcmc function on the simulated data:

data("SimulDE")
result_comp <- comp.bcmc(data.exp=SimulDE$express, data.clin=SimulDE$clin,  
                         B=5, parallel=FALSE)
names(result_comp)
head(result_comp$Rg) ## BCMC statistic
head(result_comp$pos.wp) ## predicted up-regulated weight pattern
head(result_comp$pvalue) ## permutation p-values

Note that this might take a few minutes, and we only run B=5 permutations for demonstration purpose. Examples for other two functions can be found in the R documentation.