TP53.Rmd

---
title: "TP53"
author: "almut"
date: "16 Juni 2019"
output: html_document
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

## TP53 signature

# Differentially expressed genes

### 1. Differential expression analysis
load packages
```{r packages, warning=FALSE, message=FALSE}
library(DESeq2)
library(tidyverse)
library(ggsci)
library(matrixStats)
library(piano)
library(reshape2)
library(genefilter)
library(Biobase)
library(ComplexHeatmap)
library(ggplot2)
library(gtable)
library(grid)
library(circlize)
library(gridExtra)
library(ggpubr)
library(RColorBrewer)
library(here)

```


load data
```{r datasets}
data_dir <- here("data")
output_dir <- here("output")
figure_dir <- here("output/figures")

#dds data set. gene expression data + patmetadata
load(paste0(data_dir, "/ddsrnaCLL_150218.RData"))

variant <- "TP53"
#filter for patients without NA in variant
ddsCLL <- ddsCLL[, !is.na(colData(ddsCLL)[,variant])]

#differentially expressed genes between TP53 groups (see differential expression.html)
diff_all <- read.csv(file=paste0(output_dir, "/diff_genes/", variant, "_diffGenes.csv"))


rownames(diff_all) <- diff_all$X
diff_all <- diff_all[which(diff_all$padj < 0.01 ),-1]
diff <- diff_all


mutStatus <- data.frame(colData(ddsCLL)) %>% arrange(TP53)
colnames(ddsCLL) <-colData(ddsCLL)$PatID
ddsCLL <- ddsCLL[, mutStatus$PatID]

#expression data
ddsCLL <- estimateSizeFactors(ddsCLL)
RNAnorm <- varianceStabilizingTransformation(ddsCLL, blind = T)

```


Expression matrix
```{r var gene expression}

#filter for sign. genes in variant
exprMat <- assay(RNAnorm)
exprVariant <- exprMat[rownames(diff),]
colnames(exprVariant) <- colData(ddsCLL)$PatID
exprVariant.new <- log2(exprVariant)
exprVariant.new <- t(scale(t(exprVariant.new)))
exprVariant.new[exprVariant.new > 4] <- 4
exprVariant.new[exprVariant.new < -4] <- -4
rownames(exprVariant.new) <- rowData(RNAnorm[rownames(diff),])$symbol

```


## Expression signature
```{r complex heatmap, fig.width= 22, fig.height=25}

#colors
colors = colorRamp2(c(-4,-2,0,2,4), c("#2166ac","#4393c3", "#f7f7f7", "#d6604d","#b2182b"))
annocol <- get_palette("jco", 10)
annocolor <- list(TP53 = c("1" = annocol[8], "0" = annocol[9]))
rowcolors <-colorRampPalette(brewer.pal(5, "Set1"))(5)
rowcolors[6] <- "white"


feature <- as.data.frame(colData(ddsCLL)[,c(variant)])
colnames(feature) <- c(variant)  
rownames(feature) <- colnames(RNAnorm)

ha_col <- HeatmapAnnotation(df = feature, col = annocolor, annotation_height = unit(c(rep(1.9, 1)), "cm"), annotation_legend_param = list(title_gp = gpar(fontsize = 50), labels_gp = gpar(fontsize = 45),  grid_height = unit(1.9, "cm"), grid_width = unit(1.9, "cm")))


h1 <- Heatmap(exprVariant.new ,                                                     
              km = 2,
              gap = unit(0.5, "cm"),
              cluster_columns = F,
              clustering_distance_rows = "pearson",
              clustering_method_rows = "ward.D2",
              column_title = paste0("Gene signature: ", variant),                      
              col = colors,
              column_title_gp = gpar(fontsize = 60, fontface = "bold"), 
              heatmap_legend_param = list(title = "expr", 
                                          title_gp = gpar(fontsize = 50), 
                                          grid_height = unit(1.9, "cm"), 
                                          grid_width = unit(1.9, "cm"), 
                                          gap = unit(2, "cm"), 
                                          labels_gp = gpar(fontsize = 45)), 
              column_dend_height = unit(2.5, "cm"),
              show_row_dend = FALSE, 
              show_column_names = FALSE , 
              show_row_names = TRUE, 
              row_names_gp = gpar(fontsize = 20),
              top_annotation = ha_col)


#svg(filename=paste0(figure_dir, "/", variant, "_gene_expr.svg"), width=30, height=45)
#pdf(file=paste0(figure_dir, "/", variant, "_gene_expr.pdf"), width=22, height=25)

draw(h1 ) 
#dev.off()

#draw(h1)
```


# Sample and gene specific expression - top genes

```{r single gene counts, fig.width=6, fig.height=5}
#function to create stripchart plots for specific genes
gene_count <- function(gene_nam){
  geneEnsID <- rownames(ddsCLL)[which(rowData(ddsCLL)$symbol %in% gene_nam)]
  geneNum <- counts(ddsCLL)[geneEnsID,]
  mutPat <- as.data.frame(colData(ddsCLL)[, c(variant)])
  colnames(mutPat) <- c("genotype")
  geneDat <- cbind(mutPat, geneNum)
  colnames(geneDat) <- c("genotype", "counts")
  
  p <- ggstripchart(geneDat, x = "genotype", y = "counts",
          color = "genotype",
          palette = "jco",
          add = "mean_sd",
          title = paste(gene_nam),
          font.x = 18, font.y = 18, font.legend = 16, 
          ylab = "counts") + font("xy.text", size = 15) + font("title", size = 20, face = "bold")
  #ggsave(file=paste0(figure_dir, "/tri12/genetic_interaction_", gene_nam, ".svg"), plot=p, width=6, height=5)
  p
}

diff <- diff_all[which(abs(diff_all$stat) > 4.5),]
geneList <- as.character(diff$Symbol)
geneList <- geneList[-which(geneList %in% "")]

lapply(geneList, gene_count)


```
# Interesting top genes

A list of genes, with manual checked expression differences:
Downregulated in **Del11q22**:
+ FLII
+ NUP88
+ DRG2
+ SPAG7
+ SMG6
+ RABEP1
+ ELAC2
+ TOP3A
+ MAP2K4
+ RNF167
+ PANK4
+ MED11
+ PELP1
+ MPDU1
+ CTDNEP1
+ SGF29
+ FXR2

Upregulated in **Del11q22**:
+ HYPK


# Gene set enrichment analysis

Gene sets
```{r gene sets}
#load gene set collection
#Hallmark
gsc <- loadGSC("/home/almut/Dokumente/masterarbeit/data/h.all.v6.0.symbols.gmt", type="gmt")
#Kegg
gsc_Kegg <- loadGSC("/home/almut/Dokumente/masterarbeit/data/c2.cp.kegg.v6.0.symbols.gmt", type="gmt")
```

Run piano
```{r piano, fig.width=16, fig.height=7}

gmtFile <- gsc

diff_res <- diff_all
diff_res$chromosome <- rowData(RNAnorm)[rownames(diff_res),]$chromosome
diff_res <- diff_res[-which(diff_res$Symbol %in% c("", NA)),]


geneNam <- diff_res$Symbol 
pVal <- diff_res$padj
logFold <- diff_res$log2FoldChange
stat <- diff_res$stat
gsTab <- data.frame(gene = geneNam, stat = stat, logFold = logFold)
gsaTab <- data.frame(row.names = gsTab$gene, stat = gsTab$stat)
res <- runGSA(geneLevelStats = gsaTab,
                    geneSetStat = "gsea",
                    adjMethod = "fdr", gsc=gmtFile,
                    signifMethod = "geneSampling",
                    nPerm = 50000,
                    gsSizeLim=c(1, Inf))
Res_up <- arrange(GSAsummaryTable(res), `p adj (dist.dir.up)`)            
Res_dn <- arrange(GSAsummaryTable(res), `p adj (dist.dir.dn)`)
  

#Plot
resPlot <- Res_up[, c(1:3,5,8,9)]
resPlot_dn <- Res_dn[, c(1:3,7,8,9)]
colnames(resPlot) <- c("pathway", "gene_number", "stat", "p.adj","genes_up" , "genes_dn")
colnames(resPlot_dn) <- c("pathway", "gene_number", "stat", "p.adj","genes_up" , "genes_dn")


enrichPlot <- resPlot_dn[c(1:5),] %>% mutate(log10Padj = -log10(p.adj)) 

#plot
p <- ggbarplot(enrichPlot, x = "pathway", y = "log10Padj",
          fill = "gene_number",         
          color = "white",            
          palette =  "gsea",            
          sort.val = "asc",         
          sort.by.groups = FALSE,     
          ylab = "-log10(padj)",
          legend.title = "#diff.genes",
          rotate = TRUE,
          font.x = 20, font.y = 20, font.legend = 20, legend = "right", 
          title = "TP53 - Hallmark",
          ggtheme = theme_pubr()) + 
  font("xy.text", size = 16) + 
  font("title", size = 20, face = "bold")
  
#ggsave(file=paste0(figure_dir,"/GSEA_", variant, "_Kegg.svg"), plot=p, width=14, height=7)
  p



```