cal.Activity.GS: Calculate Activity Value for Gene Sets

In jyyulab/NetBID: Network-based Bayesian Inference of Drivers, version 2

Calculate Activity Value for Gene Sets

Description

cal.Activity.GS calculates activity value for each gene set, and return a numeric matrix with rows of gene sets and columns of samples.

Usage

cal.Activity.GS(
  use_gs2gene = all_gs2gene[c("H", "CP:BIOCARTA", "CP:REACTOME", "CP:KEGG")],
  cal_mat = NULL,
  es.method = "mean",
  std = TRUE
)

Arguments

use_gs2gene	list, contains elements of gene sets. Element name is gene set name, each element contains a vector of genes belong to that gene set. Default is using all_gs2gene[c('H','CP:BIOCARTA','CP:REACTOME','CP:KEGG')], which is loaded from gs.preload.
cal_mat	numeric matrix, gene/transcript expression matrix. If want to input activity matrix, need to use 'processDriverProfile()' to pre-process the dataset. Detailed could see demo.
es.method	character, method to calculate the activity value. Users can choose from "mean", "absmean", "maxmean", "gsva", "ssgsea", "zscore" and "plage". The details for using the last four options, users can check gsva. Default is "mean".
std	logical, if TRUE, the expression matrix will be normalized by column. Default is TRUE.

Value

Return an activity matrix with rows of gene sets and columns of samples.

Examples

analysis.par <- list()
analysis.par$out.dir.DATA <- system.file('demo1','driver/DATA/',package = "NetBID2")
NetBID.loadRData(analysis.par=analysis.par,step='ms-tab')
gs.preload(use_spe='Homo sapiens',update=FALSE)
use_gs2gene <- merge_gs(all_gs2gene=all_gs2gene,
                        use_gs=c('H','CP:BIOCARTA','CP:REACTOME','CP:KEGG','C5'))
exp_mat_gene <- Biobase::exprs(analysis.par$cal.eset)
## each row is a gene symbol, if not, must convert ID first
ac_gs <- cal.Activity.GS(use_gs2gene = use_gs2gene,
                        cal_mat = exp_mat_gene)
## if want to input activity-matrix
ac_mat <- cal.Activity(target_list=analysis.par$merge.network$target_list,
          cal_mat=Biobase::exprs(analysis.par$cal.eset),
          es.method='weightedmean')
# pre-process the activity matrix by selecting the one
# with larger target size for duplicate drivers
Driver_name <- rownames(ac_mat)
ms_tab <- analysis.par$final_ms_tab
driver_size <- ms_tab[Driver_name,]$Size
use_driver <- processDriverProfile(Driver_profile=driver_size,
                              Driver_name=Driver_name,
                              choose_strategy='max',
                              return_type='driver_name')
use_driver_gene_name <-
              processDriverProfile(Driver_profile=driver_size,
                              Driver_name=Driver_name,
                              choose_strategy='max',
                              return_type='gene_name')
ac_mat_gene <- ac_mat[use_driver,]
rownames(ac_mat_gene) <- use_driver_gene_name
driver_ac_gs <- cal.Activity.GS(use_gs2gene = use_gs2gene,
                        cal_mat = ac_mat_gene)

jyyulab/NetBID documentation built on Dec. 23, 2024, 6:34 a.m.