#####RNA-seq analysis for emf33 comb libraries (OSC)
library(ggrepel)
#RNA-seq for TE only

te_class <- read.table("~/Dropbox (hannonlab)/Sequencing Data/deeptools_annotation/te_list_gl_vs_soma_merged.txt", as.is = TRUE)

te_reads <- read.table("~/Dropbox (hannonlab)/Sequencing Data/Nxf2_final/OSC/RNA-seq/comb/counting/emf33_te_reads.txt", as.is = TRUE, header = TRUE)

#Calculate RPM
rownames(te_reads) <- te_class[,1]
te_reads[,2] <- te_reads[,2]*(1000000/sum(te_reads[,2]))
te_reads[,3] <- te_reads[,3]*(1000000/sum(te_reads[,3]))
te_reads[,4] <- te_reads[,4]*(1000000/sum(te_reads[,4]))
te_reads[,5] <- te_reads[,5]*(1000000/sum(te_reads[,5]))

#Normalize to scaling factor (te/dm6 reads)
te_reads[,2] <- te_reads[,2]*0.076810813
te_reads[,3] <- te_reads[,3]*0.1851656
te_reads[,4] <- te_reads[,4]*0.114993838
te_reads[,5] <- te_reads[,5]*0.132746382


#Removes all rows with at least one value < 1 rpm
te_reads <- te_reads[!rowSums(te_reads < 1),]

# Calculate FC for all conditions treatment (eg. Nxf2) vs gfp)
fc4_te_osc <- cbind(te_reads[,3]/te_reads[,2]>4, te_reads[,4]/te_reads[,2]>4, te_reads[,5]/te_reads[,2]>4, te_reads[,4]/te_reads[,3]>4, te_reads[,5]/te_reads[,3]>4, te_reads[,5]/te_reads[,4]>4)
colnames(fc4_te_osc) <- c("piwi_gfp", "panx_gfp", "nxf2_gfp", "panx_piwi", "nxf2_piwi", "nxf2_panx")

#FC <0.25
fc4_te_osc_neg <- cbind(te_reads[,3]/te_reads[,2]<0.25, te_reads[,4]/te_reads[,2]<0.25, te_reads[,5]/te_reads[,2]<0.25, te_reads[,4]/te_reads[,3]<0.25, te_reads[,5]/te_reads[,3]<0.25, te_reads[,5]/te_reads[,4]<0.25)
colnames(fc4_te_osc_neg) <- c("piwi_gfp", "panx_gfp", "nxf2_gfp", "panx_piwi", "nxf2_piwi", "nxf2_panx")

#edit rownames and add class column
library(stringr)
test <- str_split_fixed(rownames(te_reads), "_", 2)
te_reads[,1] <- test[,1]
te_reads <- cbind(te_reads, test[,2])
colnames(te_reads) <- c("te","gfp", "piwi", "panx", "nxf2", "class")

te_reads[,2:5] <- log2(te_reads[,2:5])

library(ggplot2)


#RNA-seq for dm6

dm6_reads_gfp <- read.table("~/Dropbox (hannonlab)/Sequencing Data/Nxf2_final/OSC/RNA-seq/comb/counting/gfp_rna_.count.htseq", as.is = TRUE)
dm6_reads_piwi <- read.table("~/Dropbox (hannonlab)/Sequencing Data/Nxf2_final/OSC/RNA-seq/comb/counting/piwi_rna_.count.htseq", as.is = TRUE)
dm6_reads_panx <- read.table("~/Dropbox (hannonlab)/Sequencing Data/Nxf2_final/OSC/RNA-seq/comb/counting/panx_rna_.count.htseq", as.is = TRUE)
dm6_reads_nxf2 <- read.table("~/Dropbox (hannonlab)/Sequencing Data/Nxf2_final/OSC/RNA-seq/comb/counting/nxf2_rna_.count.htseq", as.is = TRUE)

dm6_reads <- cbind(dm6_reads_gfp, dm6_reads_piwi[,2], dm6_reads_panx[,2], dm6_reads_nxf2[,2])
dm6_reads <- dm6_reads[1:17622,]

#Calculate rpm
dm6_reads[,2] <- dm6_reads[,2]*(1000000/sum(dm6_reads[,2]))
dm6_reads[,3] <- dm6_reads[,3]*(1000000/sum(dm6_reads[,3]))
dm6_reads[,4] <- dm6_reads[,4]*(1000000/sum(dm6_reads[,4]))
dm6_reads[,5] <- dm6_reads[,5]*(1000000/sum(dm6_reads[,5]))



dm6_reads_rpm <- dm6_reads
rownames(dm6_reads_rpm) <- dm6_reads[,1]
colnames(dm6_reads_rpm) <- c("gfp", "piwi", "panx", "nxf2")
#Removes all rows with at least one value < 1
dm6_reads_rpm <- dm6_reads_rpm[!rowSums(dm6_reads_rpm < 1),]

# Calculate FC for all conditions treatment (eg. Nxf2) vs gfp)
fc4_dm6_osc <- cbind(dm6_reads_rpm[,3]/dm6_reads_rpm[,2]>4, dm6_reads_rpm[,4]/dm6_reads_rpm[,2]>4, dm6_reads_rpm[,5]/dm6_reads_rpm[,2]>4, dm6_reads_rpm[,4]/dm6_reads_rpm[,3]>4, dm6_reads_rpm[,5]/dm6_reads_rpm[,3]>4, dm6_reads_rpm[,5]/dm6_reads_rpm[,4]>4)
colnames(fc4_dm6_osc) <- c("piwi_gfp", "panx_gfp", "nxf2_gfp", "panx_piwi", "nxf2_piwi", "nxf2_panx")

fc4_dm6_osc_neg <- cbind(dm6_reads_rpm[,3]/dm6_reads_rpm[,2]<0.25, dm6_reads_rpm[,4]/dm6_reads_rpm[,2]<0.25, dm6_reads_rpm[,5]/dm6_reads_rpm[,2]<0.25, dm6_reads_rpm[,4]/dm6_reads_rpm[,3]<0.25, dm6_reads_rpm[,5]/dm6_reads_rpm[,3]<0.25, dm6_reads_rpm[,5]/dm6_reads_rpm[,4]<0.25)
colnames(fc4_dm6_osc_neg) <- c("piwi_gfp_neg", "panx_gfp_neg", "nxf2_gfp_neg", "panx_piwi_neg", "nxf2_piwi_neg", "nxf2_panx_neg")


dm6_reads_rpm_log2 <- log2(dm6_reads_rpm[2:5])


# Combine Genes and TE in one plot

colnames(te_reads) <- c("gene", "gfp", "piwi", "panx", "nxf2", "class")
dm6_reads_rpm_log2 <- cbind(rownames(dm6_reads_rpm_log2),dm6_reads_rpm_log2)
dm6_reads_rpm_log2 <- data.frame(dm6_reads_rpm_log2,"gene")
colnames(dm6_reads_rpm_log2) <- c("gene", "gfp", "piwi", "panx", "nxf2","class")
te_dm6_comb_reads_log2 <- rbind(dm6_reads_rpm_log2,te_reads[1:6])

#########Plot all upregultated TE names##########

#fc_all <- rbind(fc4_dm6_osc,fc4_te_osc)
#fc_all_neg <- rbind(fc4_dm6_osc_neg,fc4_te_osc_neg)

dummy <- as.data.frame(matrix(nrow = 7841, ncol = 6, FALSE))
colnames(dummy) <- c("piwi_gfp", "panx_gfp", "nxf2_gfp", "panx_piwi", "nxf2_piwi", "nxf2_panx")
fc_all <- rbind(dummy,fc4_te_osc)
fc_all_neg <- rbind(dummy,fc4_te_osc_neg)
te_dm6_comb_reads_log2 <- cbind(te_dm6_comb_reads_log2,fc_all,fc_all_neg)