inst/R_scripts/TCGA_liver/02.runMethods.R

In CNRGH/crimmix: Multi-omics Integration

################################################
## Run methods on Liver cancer dataset
################################################
library(CrIMMix)
library(dplyr)
library(FlowSOM)

source("inst/R_scripts/TCGA_Liver/01.loadData.R")
str(liver)
str(clinical_data)
K <- clinical_data$pathology_T_stage %>% nlevels
### Filter
## Number of data set used for the analysis
idx_dat <- c(1,2,3,4)
set.seed(33)
## Extract patients:
path_stage <- clinical_data$pathology_T_stage %>% levels
samples <- sapply(path_stage, function(p_s){
  clinical_data$id <- rownames(clinical_data)
  sub <- clinical_data %>% filter(pathology_T_stage==p_s)
  sample(sub$id, size = 0.75*nrow(sub))
}) %>% unlist


liver_filter <- lapply(liver, function(dd){
  dd[samples,]
})
true.clust_liver <- clinical_data[samples, "pathology_T_stage"]
true.clust_liver %>% table
## filter mutations
idx.genes.non.mutated <- which(liver_filter$mutation %>% colSums==0)
liver_filter$mutation <- liver_filter$mutation[, -idx.genes.non.mutated]

# SNFresults <- IntMultiOmics(liver_filter, method = "SNF", K=4)
# saveRDS(SNFresults, "inst/extdata/TCGA/Liver/snf_res.rds")
# 
# 
# c_1 <- (5/sqrt((liver_filter[idx_dat] %>% sapply(dim))[2,])) %>% round(2)
# SGCCAresults <- IntMultiOmics(liver_filter[idx_dat], method = "SGCCA", K = 4, c1 = c_1,
#                               ncomp = rep(5, length(liver_filter[idx_dat])))
# SGCCAresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# saveRDS(SGCCAresults, "inst/extdata/TCGA/Liver/sgcca_res.rds")
# 
# ## up to 0.139792 (bad)
# MoClusterresults <- IntMultiOmics(liver_filter[idx_dat], method = "Mocluster", K = 4,
#                                   k = c(0.01, 0.1, 0.01, 0.005)[idx_dat],ncomp = 4)
# saveRDS(MoClusterresults, "inst/extdata/TCGA/Liver/mocluster_res.rds")
# MoClusterresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# 
# 
# #iClusterresults <- IntMultiOmics(liver_filter, method = "iCluster", K=K-1)
# liver_filter_mcia <- liver_filter[c(1,2,4)]
# MCIArresults <- IntMultiOmics(liver_filter_mcia, method = "MCIA", K=4, ncomp=5)
# MCIArresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# saveRDS(MCIArresults, "inst/extdata/TCGA/Liver/mcia_res.rds")
# 
# 
# 
# MOFAresults <- IntMultiOmics(liver_filter[idx_dat], method = "MOFA", K=4)
# saveRDS(MOFAresults, "inst/extdata/BXObesityD/mofa_res.rds")
# 
# Kernelresults <- IntMultiOmics(liver_filter[c(1,3)], method = "MixKernel", K=4)
# Kernelresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# saveRDS(Kernelresults, "inst/extdata/TCGA/Liver/kernel_res.rds")
# 
# NFMresults <- IntMultiOmics(liver_filter[c(1,2)], method = "intNMF", K=4)
# NFMresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# saveRDS(NFMresults, "inst/extdata/TCGA/Liver/NMF_res.rds")
# 
# RGCCAresults <- IntMultiOmics(liver_filter[idx_dat], method = "RGCCA", K=4, ncomp=rep(4,  length(liver_filter[idx_dat])))
# RGCCAresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# saveRDS(RGCCAresults, "inst/extdata/TCGA/Liver/RGCCA_res.rds")
# 
# iClusterresults <- IntMultiOmics(liver_filter[idx_dat], method = "iCluster", K=3, type=c("gaussian", "binomial", "gaussian", "multinomial"))
# iClusterresults$clust %>% FMeasure(predictedClusters=true.clust_liver %>% as.numeric() , silent = FALSE)
# saveRDS(iClusterresults, "inst/extdata/TCGA/Liver/icluster_res.rds")
# 


print("CIMLR")
CIMLR_results <- liver_filter %>% IntMultiOmics(method="CIMLR", K=4)
saveRDS(CIMLR_results, file=file.path(sprintf("inst/extdata/TCGA/Liver/CIMLR_res.rds")))

print("LRAcluster")
LRAcluster_results <- liver_filter %>% IntMultiOmics(method="LRAcluster", K=4, type=c("gaussian", "binary", "gaussian","gaussian"))
saveRDS(LRAcluster_results, file=file.path(sprintf("inst/extdata/TCGA/Liver/LRAcluster_res.rds")))

print("PINSPLUS")
PINSPLUS_results <- liver_filter %>% IntMultiOmics(method="PINSPlus", K=4)
saveRDS(PINSPLUS_results, file=file.path(sprintf("inst/extdata/TCGA/Liver/PINSPLUS_res.rds")))

print("Consensus Clustering")
ConsensusClustering_results <- liver_filter %>% IntMultiOmics(method="ConsensusClustering", K=4)
saveRDS(ConsensusClustering_results, file=file.path(sprintf("inst/extdata/TCGA/Liver/ConsensusClustering_res.rds")))