generate.masterTable: Generate the Master Table for Drivers

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

Generate the Master Table for Drivers

Description

generate.masterTable generates a master table to show the mega information of all tested drivers.

Usage

generate.masterTable(
  use_comp = NULL,
  DE = NULL,
  DA = NULL,
  target_list = NULL,
  main_id_type = NULL,
  transfer_tab = NULL,
  tf_sigs = tf_sigs,
  z_col = "Z-statistics",
  display_col = c("logFC", "P.Value"),
  column_order_strategy = "type"
)

Arguments

use_comp	a vector of characters, the name of multiple comparisons. It will be used to name the columns of master table.
DE	list, a list of DE comparisons, each comparison is a data.frame. The element name in the list must contain the name in use_comp.
DA	list, a list of DA comparisons, each comparison is a data.frame. The element name in the list must contain the name in use_comp.
target_list	list, a driver-to-target list. The names of the list elements are drivers. Each element is a data frame, usually contains three columns. "target", target gene names; "MI", mutual information; "spearman", spearman correlation coefficient. It is highly suggested to follow the NetBID2 pipeline, and the TF_network could be generated by get_net2target_list and get.SJAracne.network.
main_id_type	character, the type of driver's ID. It comes from the attribute name in biomaRt package. Such as "ensembl_gene_id", "ensembl_gene_id_version", "ensembl_transcript_id", "ensembl_transcript_id_version" or "refseq_mrna". For details, user can call biomaRt::listAttributes() to display all available attributes in the selected dataset.
transfer_tab	data.frame, the data frame for ID conversion. This can be obtained by calling get_IDtransfer. If NULL and main_id_type is not in the column names of tf_sigs, it will use the conversion table within the function. Default is NULL.
tf_sigs	list, contains all the detailed information of TF and Sig. Users can call db.preload for access.
z_col	character, name of the column in DE and DA contains the Z statistics. Default is "Z-statistics".
display_col	character, name of the column in DE and DA need to be kept in the master table. Default is c("logFC","P.Value").
column_order_strategy	character, users can choose between "type" and "comp". Default is "type". If set as type, the columns will be ordered by column type; If set as comp, the columns will be ordered by comparison.

Details

The master table gathers TF (transcription factor) information, Sig (signaling factor) information, all the DE (differential expression analysis) and DA (differential activity analysis) from multiple comparisons. It also shows each driver's target gene size and other additional information (e.g. gene biotype, chromosome name, position etc.).

Value

Return a data frame contains the mega information of all tested drivers. The column "originalID" and "originalID_label" is the same ID as from the original dataset.

Examples

analysis.par <- list()
analysis.par$out.dir.DATA <- system.file('demo1','driver/DATA/',package = "NetBID2")
NetBID.loadRData(analysis.par=analysis.par,step='ms-tab')
#analysis.par$final_ms_tab ## this is master table generated before
ac_mat <- cal.Activity(target_list=analysis.par$merge.network$target_list,
                       cal_mat=Biobase::exprs(analysis.par$cal.eset),es.method='weightedmean')
analysis.par$ac.merge.eset  <- generate.eset(exp_mat=ac_mat,
                                             phenotype_info=Biobase::pData(analysis.par$cal.eset))
phe_info <- Biobase::pData(analysis.par$cal.eset)
all_subgroup <- base::unique(phe_info$subgroup) ##
for(each_subtype in all_subgroup){
 comp_name <- sprintf('%s.Vs.others',each_subtype) ## each comparison must give a name !!!
 G0  <- rownames(phe_info)[which(phe_info$`subgroup`!=each_subtype)] # get sample list for G0
 G1  <- rownames(phe_info)[which(phe_info$`subgroup`==each_subtype)] # get sample list for G1
 DE_gene_limma <- getDE.limma.2G(eset=analysis.par$cal.eset,G1=G1,G0=G0,
                                 G1_name=each_subtype,G0_name='other')
 analysis.par$DE[[comp_name]] <- DE_gene_limma
 DA_driver_limma <- getDE.limma.2G(eset=analysis.par$ac.merge.eset,G1=G1,G0=G0,
                                   G1_name=each_subtype,G0_name='other')
 analysis.par$DA[[comp_name]] <- DA_driver_limma
}
all_comp <- names(analysis.par$DE) ## get all comparison name for output
db.preload(use_level='gene',use_spe='human',update=FALSE);
test_ms_tab <- generate.masterTable(use_comp=all_comp,
                                           DE=analysis.par$DE,
                                           DA=analysis.par$DA,
                                           target_list=analysis.par$merge.network$target_list,
                                           tf_sigs=tf_sigs,
                                           z_col='Z-statistics',
                                           display_col=c('logFC','P.Value'),
                                           main_id_type='external_gene_name')

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