E14.5_Nkx2Dot1lcKO_VT_Deseq2.Rmd

---
title: "E14.5 Nkx2.1-Dot1lcKO Ventral telencephalon RNA-seq DE analysis"
output:
  html_document: default
  pdf_document: default
date: '2022-10-04'
---

```{r message=FALSE, warning=FALSE, include=FALSE}
library(knitr)
# Set so that long lines in R will be wrapped:
opts_chunk$set(tidy.opts=list(width.cutoff=80), tidy=TRUE)
options(width = 80)
```

load packages
```{r message=FALSE, warning=FALSE, echo=FALSE}
library(DESeq2)
library(limma)
library(clusterProfiler)
library(org.Mm.eg.db)
library(EnhancedVolcano)

library(reshape2)
library(scales)
library(plyr)
library(pheatmap)
```

| Dataset name | genotype | sex    | litter | Sample name |
|--------------|----------|--------|--------|-------------|
| KO1          | KO       | MALE   | 3      | VT_7        |
| KO2          | KO       | FEMALE | 2      | VT_6        |
| WT1          | WT       | FEMALE | 2      | VT_5        |
| WT2          | WT       | FEMALE | 2      | VT_4        |
| WT3          | WT       | MALE   | 1      | VT_2        |
| KO3          | KO       | FEMALE | 1      | VT_3        |
| WT4          | WT       | MALE   | 1      | VT_1        |
| WT5          | WT       | MALE   | 3      | VT_10       |
| KO4          | KO       | FEMALE | 3      | VT_8        |
| KO5          | KO       | MALE   | 3      | VT_9        |

: Nkx2.1_Dot1lcKO E14.5 ventral telencephalon (VT) RNA-seq sample metadata

------------------------------------------------------------------------

load raw count matrix from featurecounts (Galaxy) and analysis design file
```{r}
nkxdot.featurecount<-read.table(
    "C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/Galaxy338-[nkx2.1_Dot1lcKO_featurecount_matrix].tabular", 
    header=TRUE, 
    sep="\t", 
    quote="",
    fill=TRUE, 
    row.names = "Geneid") 
#this is from running featurecounts on each dataset and joining them in one dataset
#featurecounts parameters: 
# -s  0  -Q  10  -t 'exon' -g 'gene_id'  --minOverlap  1 --fracOverlap 0 --fracOverlapFeature 0 -p  -C 

nkxdot.info<-read.table("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/nkx2dot1lcko_deseq2_design.txt",
    header=TRUE, 
    sep="\t", 
    quote="", 
    fill=FALSE, 
    row.names = 1) 

# this is just a text file generated in notepad specifying properties of the datasets
#metadata
nkxdot.info$sex<-factor(nkxdot.info$sex) 
nkxdot.info$genotype<-factor(nkxdot.info$genotype)
nkxdot.info$litter<-factor(nkxdot.info$litter) 
#sex, litter and genotype columns were character class 
#this converts them to factors (what DESEq2 needs).
```
Create DESeq2 dataset from the above input
```{r}
nkxdot.featurecount <- nkxdot.featurecount[, rownames(nkxdot.info)] 
all(rownames(nkxdot.info) == colnames(nkxdot.featurecount)) 
#this makes sure rownames of the design are the same as column names of the count matrix.

nkxdot.dds <- DESeqDataSetFromMatrix(countData = nkxdot.featurecount,
                                     colData = nkxdot.info,
                                     design = ~ sex + litter + genotype) 
#this generates the Deseq2 dataset (dds), which will be used for PCA plots, 
#count normalisation,batch removal and DE analysis, etc. simply for everything downstream :) 

nkxdot.dds
```

Normalise counts (VST) and create PCA plot to explore data
```{r}
vsd<-vst(nkxdot.dds) 
#count normalisation (Variance stabilising transformation; VST)
nkxdot_beforecorrection_PCA<-plotPCA(vsd,
                                    intgroup=c("genotype", "sex", "litter"))

print(nkxdot_beforecorrection_PCA)
dev.off()
nkxdot_beforecorrection_PCA
#plot PCA and label the datasets according to genotype, sex and litter
```

-   As can be seen from the PCA plot, the datasets separate mainly by sex (male vs. female; PC1) rather than the genotype. There is a strong sex and litter (batch) bias in the data. We need to see the data when sex and litter effects are removed.

use limma to do batch correction
```{r}
mat <- assay(vsd)
#create a matrix from VST-normalised counts for limma
mm <- model.matrix(~genotype, colData(vsd)) 
#define the model for limma, her we chose ~genotype, because that is the main factor we are interested in.
mat <- limma::removeBatchEffect(mat, batch=vsd$sex, batch2= vsd$litter, design=mm) 
#remove the batch effect (variation) coming from sex and litter of the samples. this is saved over the initial matrix (mat).
#this is the normalised batch-corrected counts now.
counts.after.batchcorrection<-mat
#rename mat to something more intuitive for later count table
assay(vsd) <- mat 
#rewrite mat to run plotPCA on it

nkxdot_aftercorrection_PCA<-plotPCA(vsd,
                                    intgroup=c("genotype", "sex", "litter")) 
pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/nkxdot_aftercorrection_PCA.pdf",
    width=4, height=5)
print(nkxdot_aftercorrection_PCA)
dev.off()

#PCA plot with the sex and litter variances removed.
```

-   Now the sex and litter biases are gone. However, this reveals some undesired variation in WT:m:1 (purple) dataset that clusters with KO datasets. Removing this dataset is justified (OUTLIER).

Remove #WT4 (outlier) from the count matrix and design file
```{r}
#remove WT4
nkxdot.featurecount.rm <- as.matrix((nkxdot.featurecount[,-7]),
                                    sep="\t", 
                                    row.names="Geneid") 
#remove info of WT4
nkxdot.info.rm<-nkxdot.info[-7,] 
nkxdot.featurecount.rm <- nkxdot.featurecount.rm[, rownames(nkxdot.info.rm)]
#create a new deseq2 dataset without WT4 dataset. rm for "dataset removed"
all(rownames(nkxdot.info.rm) == colnames(nkxdot.featurecount.rm))
nkxdot.rm.dds <- DESeqDataSetFromMatrix(countData = nkxdot.featurecount.rm,
                                        colData = nkxdot.info.rm,
                                        design = ~ sex + litter + genotype)
nkxdot.rm.dds
```

re-plot the PCA with the new count matrix (rm) after batch correction
```{r}
vsd.rm<-vst(nkxdot.rm.dds)
mat.rm <- assay(vsd.rm) 
#create a matrix from VST-normalised counts for limma
mm.rm <- model.matrix(~genotype, 
                      colData(vsd.rm)) 
#define the model for limma, her we chose ~genotype, because that is the main factor we are interested in.
mat.rm <- limma::removeBatchEffect(mat.rm, 
                                   batch=vsd.rm$sex, 
                                   batch2= vsd.rm$litter, 
                                   design=mm.rm) 
#remove the batch effect (variation) coming from sex and litter of the samples. this is saved over the initial matrix (mat).this is the normalised batch-corrected counts now.
counts.after.batchcorrection<-mat.rm 
#rename mat to something more intuitive for later count table
assay(vsd.rm) <- mat.rm 
#rewrite mat to run plotPCA on it

plotPCA(vsd.rm, intgroup=c("genotype", "sex", "litter")) 
#PCA plot with the sex and litter variances removed.
```

-   With the WT4 dataset removed, there is better separation of WT and KO clusters.
-   There is still more variability among KO datasets.
-   Now the dataset is ready to run DESEQ2 Differential expression analysis.

Save the VST-normalised counts without WT4 and batch effects
```{r}
entrezid.counts<-row.names(counts.after.batchcorrection) #get ENTREZIDs
rownames(counts.after.batchcorrection)<-NULL #remove rownames
counts.after.batchcorrection.df<-cbind(entrezid.counts,
                                       counts.after.batchcorrection) 
#join ENTREZIDs and the count table so that the entrezids are another column
counts.after.batchcorrection.df<-as.data.frame(counts.after.batchcorrection.df) 
nkxdot.counts.after.batchcorrection.symbol<-bitr(counts.after.batchcorrection.df$entrezid.counts, 
                                                 fromType = "ENTREZID",
                                                 toType = "SYMBOL",
                                                 OrgDb = org.Mm.eg.db,
                                                 drop = FALSE) #translate ENTREZID to GENE SYMBOL

nkxdot.counts.after.batchcorrection.symbol.df<-cbind(nkxdot.counts.after.batchcorrection.symbol, 
                                                     counts.after.batchcorrection.df) 
#join GENE SYMBOLs with the count table

write.table(nkxdot.counts.after.batchcorrection.symbol.df, 
            file = "C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_vst_counts_batchcorrected.txt",
            sep = '\t', 
            quote = FALSE,)
#save the count table as tab-deliminated file
```

Run differential expression (DE) analysis with DESEQ2
```{r}
nkxdot.deseq<-DESeq(nkxdot.rm.dds) #DE analysis

res.nkxdot.vt<- results(nkxdot.deseq,
                        contrast = c("genotype", "KO", "WT")) 
#specify the parameters of the results. We want the DE genes depending on genotype between KO vs. WT.
res.nkxdot.vt.df<-as.data.frame(res.nkxdot.vt) 
#this gives the DE table with log2FC and padj etc values.

entrezid<-row.names(res.nkxdot.vt.df)
rownames(res.nkxdot.vt.df)<-NULL
res.nkxdot.vt.df<-cbind(entrezid, 
                        res.nkxdot.vt.df)
# as before, creating a entrezid column in the DE results table

nkxdot.deseq.symbol<-bitr(res.nkxdot.vt.df$entrezid,
                          fromType = "ENTREZID",
                          toType = "SYMBOL",
                          OrgDb = org.Mm.eg.db,
                          drop = FALSE) 
# translate entrezids of DEGs to gene symbol

res.nkxdot.vt.df<-cbind(nkxdot.deseq.symbol, 
                        res.nkxdot.vt.df) 
# add the gene symbols to the DE table
res.nkxdot.vt.df_sorted = res.nkxdot.vt.df[order(res.nkxdot.vt.df$padj),]
# sort the DE table by ascending padj values
write.table(res.nkxdot.vt.df_sorted, 
            file="C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_batchcorrected_deseq2_symbol_padjsorted.txt", 
            sep = "\t",
            quote = FALSE,)

res.nkxdot.vt.df_padj0.1 = res.nkxdot.vt.df[(res.nkxdot.vt.df$padj<=0.1) & 
                                                (is.na(res.nkxdot.vt.df$padj)==FALSE),]
# filter the DE table to include the genes with padj value smaller than 0.1
write.table(res.nkxdot.vt.df_padj0.1,
            file="C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_batchcorrected_deseq2_symbol_padjcutoff.txt",
            sep = "\t",
            quote = FALSE,)

res.nkxdot.vt.shrink<-lfcShrink(nkxdot.deseq, 
                                contrast = c("genotype", "KO", "WT"), 
                                res = res.nkxdot.vt, type ="normal" ) # Shrink the log2FC (LFC) values
res.nkxdot.vt.shrink.df<-as.data.frame(res.nkxdot.vt.shrink)

entrezid<-row.names(res.nkxdot.vt.shrink.df)
rownames(res.nkxdot.vt.shrink.df)<-NULL
res.nkxdot.vt.shrink.df<-cbind(entrezid, res.nkxdot.vt.shrink.df)

nkxdot.deseq.shrink.symbol<-bitr(res.nkxdot.vt.shrink.df$entrezid, 
                                 fromType = "ENTREZID",
                                 toType = "SYMBOL",
                                 OrgDb = org.Mm.eg.db,
                                 drop = FALSE) # translate entrezids of DEGs to gene symbol

res.nkxdot.vt.shrink.df<-cbind(nkxdot.deseq.shrink.symbol,
                               res.nkxdot.vt.shrink.df)
res.nkxdot.vt.shrink.df<-res.nkxdot.vt.shrink.df[(is.na(res.nkxdot.vt.shrink.df$padj)==FALSE),]
res.nkxdot.vt.shrink.df_sorted = res.nkxdot.vt.shrink.df[order(res.nkxdot.vt.shrink.df$padj),] 
# sort the LFC shrunk DE table by ascending padj values
write.table(res.nkxdot.vt.shrink.df_sorted,
            file="C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_batchcorrected_deseq2_shrink_symbol_padjsorted.txt", 
            sep = "\t", 
            quote = FALSE,)

res.nkxdot.vt.shrink.df_padj0.1 = res.nkxdot.vt.shrink.df[(res.nkxdot.vt.shrink.df$padj<=0.1) & 
                                                              (is.na(res.nkxdot.vt.shrink.df$padj)==FALSE),]  
# filter the DE table to include the genes with padj value smaller than 0.1
write.table(res.nkxdot.vt.shrink.df_padj0.1,
            file="C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_batchcorrected_deseq2_shrink_symbol_padjcutoff.txt", 
            sep = "\t", 
            quote = FALSE,)

# combine normalised counts with deseq table to filter the count table by padj<0.01
nkxdot.norm.counts.padj01<-merge(res.nkxdot.vt.df_padj0.1, 
                                 counts.after.batchcorrection.df,
                                 by.x="ENTREZID",
                                 by.y="entrezid.counts")
#filter padj<0.05
nkxdot.norm.counts.padj005 = nkxdot.norm.counts.padj01[(nkxdot.norm.counts.padj01$padj<=0.05) & 
                                                    (is.na(nkxdot.norm.counts.padj01$padj)==FALSE),]  


write.table(nkxdot.norm.counts.padj01,
            file="C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_batchcorrected_norm_counts_padjcutoff.txt", 
            sep = "\t", 
            quote = FALSE,)

```

------------------------------------------------------------------------

Volcano plot of DEGs from E14.5 Nkx2.1 Dot1lKO VT samples

```{r, fig.width=6, fig.height=6}
#get rid of NA padj values
res.nkxdot.vt.shrink.df<-res.nkxdot.vt.shrink.df[(is.na(res.nkxdot.vt.shrink.df$SYMBOL)==FALSE),] 
 #get rid of NA gene symbols
row.names(res.nkxdot.vt.shrink.df) <- make.names(res.nkxdot.vt.shrink.df[,"SYMBOL"],TRUE)

#color key
keyvals_nkx<- ifelse(
    abs(res.nkxdot.vt.shrink.df$log2FoldChange) > 0 & 
        res.nkxdot.vt.shrink.df$padj<=0.1, 'indianred4',
    ifelse('grey'))
keyvals_nkx[is.na(keyvals_nkx)] <- 'grey'
names(keyvals_nkx)[keyvals_nkx == 'indianred4'] <- 'mid'
names(keyvals_nkx)[keyvals_nkx == 'grey'] <- 'low' 

pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkx2_dot1lcko_vt_deseq2_shrink_volcanoplot.pdf", 
    width=5,
    height=7,
    compress = TRUE)

volcano_plot_nkxdot<- EnhancedVolcano(res.nkxdot.vt.shrink.df,
                lab = rownames(res.nkxdot.vt.shrink.df), 
                x = "log2FoldChange",
                y = "padj",
                xlab = bquote(~Log[2]~ 'fold change'),
                ylab = bquote(~-Log[10]~italic(P)),
                selectLab = rownames(res.nkxdot.vt.shrink.df)[which(names(keyvals_nkx) %in% c('mid'))],
                axisLabSize = 12,
                labSize = 3,
                labCol = 'grey12',
                labFace = 'plain',
                boxedLabels = FALSE,
                pointSize= 2.5,
                pCutoff = 0.1,
                FCcutoff = 0.58,
                xlim=c(-0.8, 0.8),
                ylim=c(0,-log10(10e-11)),
                colAlpha = 1/2,
                col = c("slategray", "slategray", "indianred", "dodgerblue4"),
                                    colGradient = NULL,
                                        shadeAlpha = 1/2,
                                      shadeFill = 'grey',
                                      shadeSize = 0.01,
                                      shadeBins = 2,
                                      vlineType = 'dashed',
                                      vlineCol = 'black',
                                      vlineWidth = 0.4,
                                      hline = c(0.1),
                                      hlineType = 'longdash',
                                      hlineCol = 'black',
                                      hlineWidth = 0.4,
                                      gridlines.major = TRUE,
                                      gridlines.minor = FALSE,
                                      border = "full",
                                      borderWidth = 0.8,
                                      borderColour = "black",
                                      drawConnectors=FALSE,
                                      widthConnectors = 0.5)
                print(volcano_plot_nkxdot)
dev.off()
volcano_plot_nkxdot
```

GO-term enrichment analysis
```{r, fig.height=10, fig.width=6}
#filter up- and down-regulated genes

res.nkxdot.vt.shrink.df.up<- res.nkxdot.vt.shrink.df[(res.nkxdot.vt.shrink.df$log2FoldChange> 0 & 
                                                          res.nkxdot.vt.shrink.df$padj<=0.05),]
res.nkxdot.vt.shrink.df.down<- res.nkxdot.vt.shrink.df[(res.nkxdot.vt.shrink.df$log2FoldChange<0 & 
                                                            res.nkxdot.vt.shrink.df$padj<=0.05),]

# make a list of increased and decreased genes
nkxdot.list<-list(increase=res.nkxdot.vt.shrink.df.up$ENTREZID,
                  decrease=res.nkxdot.vt.shrink.df.down$ENTREZID)

# run differential GO-term enrichment analysis

nkxdot.compare<- compareCluster(geneClusters = nkxdot.list,
                                fun = "enrichGO",
                                OrgDb = "org.Mm.eg.db",
                                keyType = "ENTREZID",
                                ont = "BP",
                                pAdjustMethod = "BH",
                                qvalueCutoff = 0.05,
                                pvalueCutoff = 0.05,
                                readable = TRUE)
# simplify terms to avoid redundancy
nkxdot.compare.simp<-simplify(nkxdot.compare,
         cutoff = 0.5,
         by = "p.adjust",
         select_fun = min,
         measure = "Wang",
         semData = NULL)
# export the dataframe of GO enrichment results
nkxdot.compare.df<-as.data.frame(nkxdot.compare)
write.table(nkxdot.compare.df, 
            file = "C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_GOterms.txt",
            sep = "\t", 
            quote = FALSE,)

# export the dataframe of GO enrichment results (simplified)
nkxdot.compare.simp.df<-as.data.frame(nkxdot.compare.simp)
write.table(nkxdot.compare.simp.df, 
            file = "C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_GOterms_simplified.txt",
            sep = "\t",
            quote = FALSE,)

# Dotplot of differential GO terms 

dp.nkxdot = dotplot(nkxdot.compare,
             showCategory=12,
             font.size=10,
             )

# Dotplot of differential GO terms simplified
dp.nkxdot.simp = dotplot(nkxdot.compare.simp,
             showCategory=12,
             font.size=10,
             )
```
GO term enrichment analysis of molecular functions
```{r, fig.height=10, fig.width=6}
#run differential GO-term enrichment analysis with molecular functions (MF)
nkxdot.compare.mf<- compareCluster(geneClusters = nkxdot.list,
                                fun = "enrichGO",
                                OrgDb = "org.Mm.eg.db",
                                keyType = "ENTREZID",
                                ont = "MF",
                                pAdjustMethod = "BH",
                                qvalueCutoff = 0.05,
                                pvalueCutoff = 0.05,
                                readable = TRUE)
#simplify terms to avoid redundancy
nkxdot.compare.mf.simp<-simplify(nkxdot.compare.mf,
         cutoff = 0.5,
         by = "p.adjust",
         select_fun = min,
         measure = "Wang",
         semData = NULL)

# export the dataframe of GO enrichment results (MF)
nkxdot.compare.mf.df<-as.data.frame(nkxdot.compare.mf)
write.table(nkxdot.compare.mf.df, 
            file = "C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_GOterms_MF.txt",
            sep = "\t",
            quote = FALSE,)

# export the dataframe of GO enrichment results (MF, simplified)
nkxdot.compare.mf.simp.df<-as.data.frame(nkxdot.compare.mf.simp)
write.table(nkxdot.compare.mf.simp.df, 
            file = "C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_GOterms_MF_simplified.txt",
            sep = "\t", 
            quote = FALSE,)

# Dotplot of differential GO terms MF

dp.nkxdot.mf = dotplot(nkxdot.compare.mf,
             showCategory=12,
             font.size=10,
             )

# Dotplot of differential GO terms MF simplified
dp.nkxdot.mf.simp = dotplot(nkxdot.compare.mf.simp,
             showCategory=12,
             font.size=10,
             )
```
export the plots
```{r}
pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/160123_nkxdot_diff_GO_analysis.pdf",
    width=5,
    height=8)
print(dp.nkxdot)
dev.off()
pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_diff_GO_analysis_MF.pdf",
    width=5, 
    height=10)
print(dp.nkxdot.mf)
dev.off()
pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/160123_nkxdot_diff_GO_analysis_simp.pdf", 
    width=5, 
    height=7)
print(dp.nkxdot.simp)
dev.off()
pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_diff_GO_analysis_MF_simp.pdf",
    width=5,
    height=10)
print(dp.nkxdot.mf.simp)
dev.off()
```
### heatmap of norm counts of DEGs with padj<=0.05
```{r, fig.align='center', fig.height=9, fig.width=5}
nkxdot.norm.counts.padj005.df<- as.data.frame(nkxdot.norm.counts.padj005[,-(4:9)])
nkxdot.norm.counts.padj005.df<- as.data.frame(nkxdot.norm.counts.padj005.df[,-c(1:2)])
nkxdot.norm.counts.padj005.df<- nkxdot.norm.counts.padj005.df[!duplicated(nkxdot.norm.counts.padj005.df),]
nkxdot.norm.counts.padj005.df$KO1<-as.numeric(nkxdot.norm.counts.padj005.df$KO1)
nkxdot.norm.counts.padj005.df$KO2<-as.numeric(nkxdot.norm.counts.padj005.df$KO2)
nkxdot.norm.counts.padj005.df$KO3<-as.numeric(nkxdot.norm.counts.padj005.df$KO3)
nkxdot.norm.counts.padj005.df$KO4<-as.numeric(nkxdot.norm.counts.padj005.df$KO4)
nkxdot.norm.counts.padj005.df$KO5<-as.numeric(nkxdot.norm.counts.padj005.df$KO5)
nkxdot.norm.counts.padj005.df$WT1<-as.numeric(nkxdot.norm.counts.padj005.df$WT1)
nkxdot.norm.counts.padj005.df$WT2<-as.numeric(nkxdot.norm.counts.padj005.df$WT2)
nkxdot.norm.counts.padj005.df$WT3<-as.numeric(nkxdot.norm.counts.padj005.df$WT3)
nkxdot.norm.counts.padj005.df$WT5<-as.numeric(nkxdot.norm.counts.padj005.df$WT5)

rownames(nkxdot.norm.counts.padj005.df) <-nkxdot.norm.counts.padj005.df$entrezid
nkxdot.norm.counts.padj005.df<- as.matrix(nkxdot.norm.counts.padj005.df[,-1])
nkxdot.norm.counts.padj005.df.heatmap<- pheatmap(nkxdot.norm.counts.padj005.df,cluster_cols = T, 
                                       border_color = "white", 
                                       clustering_distance_rows = "euclidean", 
                                       color = hcl.colors(80, "RdBu", rev=TRUE, fixup = T),
                                       angle_col = 0, 
                                       fontsize = 8, 
                                       show_colnames = T, 
                                       show_rownames = F,
                                       scale = "row"
                                        )
pdf("C:/Users/ia1014/Downloads/nkx2creDot1lcKO_rnaseq/301222_nkxdot_vt_scaled_count_heatmap_padj01.pdf", 
    width=5, 
    height=8)
print(nkxdot.norm.counts.padj005.df.heatmap)
dev.off()
```

```{r}
sessionInfo()
```