# Nacteni dat
load("ER_vysledky.rdata")
load("GSE20194_MDACC_Expression.rdata")
clinical<-read.csv("GSE20194_MDACC_Sample_Info.csv", header=T)

# kontrola
head(ER_vysledky)
head(X)
head(clinical)

#načtení balíku fgsea

library(fgsea)

#načtení balíku pro správu msigdb genových sad
library(msigdbr) #více info https://cran.r-project.org/web/packages/msigdbr/vignettes/msigdbr-intro.html

#vybereme napr. hallmark pathways
h_gene_sets = msigdbr(species = "human", category = "H")

#pripravime k analyze
h_list_GS = split(x = h_gene_sets$gene_symbol, f = h_gene_sets$gs_name)

# připravíme data pro analýzu - seřadíme dle statistiky

v<-ER_vysledky$Tstat
names(v)<-ER_vysledky$Gen

#otestujeme hallmark pathways

fgseaRes <- fgseaSimple(h_list_GS, v, nperm=10000, maxSize=500)
fgseaRes$Enrichment = ifelse(fgseaRes$NES > 0 & fgseaRes$padj<0.1, "Up-regulated", ifelse(fgseaRes$NES < 0 & fgseaRes$padj<0.1, "Down-regulated","Not significant"))

View(fgseaRes)

#vizualizujeme
library(ggplot2)

filtRes=fgseaRes[which(fgseaRes$Enrichment!="Not significant"),]

total_up = sum(fgseaRes$Enrichment == "Up-regulated")
total_down = sum(fgseaRes$Enrichment == "Down-regulated")

header = paste0("Significant pathways (Up=", total_up,", Down=",    total_down, ")")

colos = setNames(c("firebrick2", "dodgerblue2"),
                 c("Up-regulated", "Down-regulated"))

g1= ggplot(filtRes, aes(reorder(pathway, NES), NES)) +
  geom_point( aes(fill = Enrichment, size = size), shape=21) +
  scale_fill_manual(values = colos ) +
  scale_size_continuous(range = c(2,10)) +
  geom_hline(yintercept = 0) +
  coord_flip() +
  labs(x="Pathway", y="Normalized Enrichment Score",
       title=header) #+ th #???

g1