R/params.R

In mbanf/MERIT: MastEr Regulator Inference Tool

# parameters to work with 

n.cpus <- 5

b.run_grn_inference = TRUE
b.run_DNA_binding_inference = TRUE
b.run_grn_stability_selection = FALSE


# genome specific agct distribution 
bg.genome <- readRDS("datasets/dna_binding_motifs/genomic.acgt.rds")

b.precompute_and_store_diffExp_and_foldchange_matrices <- FALSE
s.timeStamp.DE = "0502" # time stamp differential expression
s.timeStamp.GRN = "0502" # time stamp GRN inference


th.diffexp = 0.05
th.lead_grn_method = 0.95
th.support_grn_methods = 0.95
th.pval.known_motifs = 0.05
th.pval.treatment = 0.05 
th.pval.tissue = 0.05
th.pval = 0.05 # heatmaps - master regulator - everythong else
th.min.score.motif <- "80%"



# condition specific info ...
# v.th.diffexp <- c(0.001, 0.005, 0.01, 0.025, 0.05, 0.075, 0.1, 0.15) 
v.th.diffexp <- c(0.025, 0.05, 0.075) 


v.th.pval.tissue <- c(0.05)
v.th.pval.treatment <- c(0.05)
v.th.lead_grn_method  = 0.95
v.th.support_grn_methods = 0.95
v.th.pval.motif_binding <- c(0.05) 
s.multipleTestCorrection= "none"  # "holm"

n.mf_shuffles <- 3
b.load = FALSE
# top 1 percentile - probability of 0.99 
# eta^2 p value cutoff - after bonferroni 
th.pval.heatmaps <- 0.05
th.pval.masterRegulators <- 0.05
# n.sim <- 10 # number of background simulations
th.min.samples <- 1
th.min_number_targets = 3
th.min_number_MR_targets = 2
# b.load_mc_sim <- TRUE
b.load_results <- TRUE
b.load_treatmentfilter <- TRUE
b.paperResults <- TRUE
b.model_comparison <- FALSE
n.min_hit_links = 3


n.trees = 5000
n.trees_support = 1000
nbootstrap = 100
nstepsLARS = 5
n.mf_shuffles = 3


# set to right folder for hierarchies
#files = c("~/Documents/JunkDNA.ai/Projects/MERIT/figures/Analysis_per_condition/masterRegulatorHierarchies/Enzymes/",
 #         "~/Documents/JunkDNA.ai/Projects/MERIT/figures/Analysis_per_condition/masterRegulatorHierarchies/Domains/")


files = c("A:/junkDNA.ai/Merit/figures/Analysis_per_condition/masterRegulatorHierarchies/Enzymes/",
          "A:/junkDNA.ai/Merit/figures/Analysis_per_condition/masterRegulatorHierarchies/Domains/")


filename_tmp_motif <- paste("results/tmp_files/l.m.motifNet_RF_GRN.pval__th.min.score.motif_", th.min.score.motif,"__th.RF5000_one_CLR_LR_ES__th.motif_binding_promoter_1000kb_to_200kb.rds", sep ="")
filename_motif_pval <- paste("results/motif_binding_inference/m.motifNet_RF_GRN.pval__th.min.score.motif_", th.min.score.motif,"__th.RF5000_one_CLR_LR_ES__th.motif_binding_promoter_1000kb_to_200kb.rds", sep ="")
filename_motif_tgs  <- paste("results/motif_binding_inference/m.motifNet_RF_GRN.tgs__th.min.score.motif_", th.min.score.motif, "__th.RF5000_one_CLR_LR_ES__th.motif_binding_promoter_1000kb_to_200kb.rds", sep ="")



# set paths to datasets 

# gene expression matrix - as tab separated text file - rows genes, columns conditions
filename.geneExpression <- "datasets/gene_expression/GSE69995_re-analyzed_data_matrix.txt"
# gene expression annotation 
filename.annotation <- "datasets/gene_expression/geneExpressionAnnotation_1120.txt" # "gene_expression/geneExpressionMeta.csv"
# meta condition groups 
filename.conditionGroups <- "datasets/gene_expression/treatments.csv"
# dna binding data (!check customized sub directories)
filename.tf_binding <- "../MERIT/datasets/tf_binding/"



df.parameter_sets <- c()
for(i in 1:length(v.th.pval.motif_binding)){
  df.parameter_sets <- rbind(df.parameter_sets, 
                             data.frame(th.diffexp = expand.grid(v.th.diffexp, v.th.lead_grn_method)[,1],
                                        th.lead_grn_method= expand.grid(v.th.diffexp, v.th.lead_grn_method)[,2], 
                                        th.pval.motif_binding  = rep(v.th.pval.motif_binding[i], nrow(expand.grid(v.th.diffexp, v.th.lead_grn_method)))))
}

df.parameter_sets.complete <- c()
for(j in 1:length(v.th.pval.tissue)){
  df.parameter_sets["th.pval.tissue"] <- v.th.pval.tissue[j]
  df.parameter_sets.complete <- rbind(df.parameter_sets.complete, df.parameter_sets)
}

df.parameter_sets <- df.parameter_sets.complete
df.parameter_sets.complete <- c()
for(j in 1:length(v.th.pval.treatment)){
  df.parameter_sets["th.pval.treatment"] <- v.th.pval.treatment[j]
  df.parameter_sets.complete <- rbind(df.parameter_sets.complete, df.parameter_sets)
}
df.parameter_sets <- df.parameter_sets.complete
df.parameter_sets.complete <- c()
for(j in 1:length(v.th.support_grn_methods)){
  df.parameter_sets["th.support_grn_methods"] <- v.th.support_grn_methods[j]
  df.parameter_sets.complete <- rbind(df.parameter_sets.complete, df.parameter_sets)
}
df.parameter_sets <- df.parameter_sets.complete