FigureS4.Rmd

---
title: "Figure4"
author: "Mike"
---
#For plotting Figure S4
Load up the required functions and libraries 
```{r, echo=FALSE, results="hide"}
library(here)
library(dplyr)
library(ggplot2)
library(reshape2)
library(NMF)                # aheatmap()
library(nat)
library(elmr)
library(nat.templatebrains)
library(nat.flybrains)
download_jefferislab_registrations()
readin<-function(file) {
  as.matrix(read.table(file, header=TRUE,row.names=1))
} #For reading in the text files containing LM overlap data 
set3d <- function(pos = c("front", "left", "back", "right",
                          "ventral", "dorsal"), zoom = 0.7, ...) {
  pos <- match.arg(pos)
  m <- diag(c(1, -1, -1, 1)) # front
  if (pos == "left") {
    m <- diag(c(0, -1, 0, 1))
    m[1, 3] <- 1
    m[3, 1] <- 1
  }
  if (pos == "back") {
    m <- diag(c(-1, -1, 1, 1))
  }
  if (pos == "right") {
    m <- diag(c(0, -1, 0, 1))
    m[1, 3] <- m[3, 1] <- -1
  }
  if (pos == "ventral") {
    m <- diag(c(1, -1, -1, 1))
  }
  if (pos == "dorsal") {
    m <- diag(c(1, 1, 1, 1))
  }
  4
  view3d(userMatrix = m, zoom = zoom, ...)
}
here()
file.symlink("/Users/dolanm/Google Drive/SplitGAL4_Paper/Figures/FigureS4", here())
```

Plot the overlap matrix of LHON axons by neuropil regions as a heatmap 
```{r, echo=FALSE}
LHON_neuropil<-readin("FigureS4/Output_axonmemb_x_nc82_nc82.txt")
NAmaster<-read.xlsx(file="/Users/dolanm/Dropbox/JFRCvisitorProject/Neuroanatomy_Master.xlsx", sheetIndex = 1)
LHnts<-select(NAmaster,LHClusters., Neurotransmitter)
LHnewnames<-select(NAmaster,LHClusters., FinalNames)

#Define the colours for NT annotations 
transmitter_col<-c("red", "dodgerblue", "forestgreen", "dark magenta", "gold")
names(transmitter_col)<-c("Acetylcholine", "GABA", "Glutamate", "Acetylcholine, GABA",  "Acetylcholine, Glutamate" )

#The types need to be in the same order as the dimnames in the matrix to be read in correctly. Need to match
#up these two dataframes. 
LHONs<-dimnames(LHON_neuropil)[[1]] #Pull in the cell-types in question and get their order 
LHONs<-data.frame(Type=LHONs)
LHnts$LHClusters.<-paste0("CellType_", LHnts$LHClusters.) #Add CellType to all the names to match them with left_join, remove this with new naming system 
names(LHnts)<-c("Type", "Neurotransmitter")
#Now join the dimnames with the annotation::cell-types to find matches but maintain matrix order.
#Note you just feed aheatmap the ordered annotations, not the annotation::cell-types pairs.
#Use different columns to add more tracks 
LHONtransmitter<-left_join(x = LHONs, y = LHnts, "Type")
LHONtransmitter<-data.frame(Neurotransmitter=as.character(LHONtransmitter$Neurotransmitter))

#Change the names of LH anatomy groups to their cell-types 
oldnames<-lapply(X = dimnames(LHON_neuropil), FUN = gsub, pattern = "CellType_", replacement = "", fixed = TRUE)
oldname1<-data.frame(oldnames=oldnames[[1]]) #For the row names
oldname1<-merge(x=oldname1, by.x ="oldnames", y=LHnewnames, by.y = "LHClusters.", sort =FALSE)
newname1<-as.character(oldname1$FinalNames)

dimnames(LHON_neuropil)[[1]]<-newname1
aheatmap(x = LHON_neuropil, filename=paste0("FigureS4/LHON_neuropil.pdf"), annRow = LHONtransmitter, annColors  = list(transmitter_col)
       , fontsize = 10, Colv = FALSE,  cexRow = 2, cexCol = 5, treeheight = 10, col="black", annLegend = FALSE)
```
Determine if there is a correlation between axon overlap and dendrite overlap 
```{r, echo=FALSE}
LHONaxon<-melt(readin("FigureS4/Output_axonmemb_x_Output_axonmemb.txt"))
LHONden<-melt(readin("FigureS4/Output_den_x_Output_den.txt"))
LHONaxon$Combo<-paste0(LHONaxon$Var1, LHONaxon$Var2)
LHONaxon<-select(LHONaxon, Combo, value)
LHONden$Combo<-paste0(LHONden$Var1, LHONden$Var2)
LHONden<-select(LHONden, Combo, value)
#Merge the two dfs into one and remove the 100% match for each cell-type onto itself 
names(LHONden)<-c("Combo", "den_value")
names(LHONaxon)<-c("Combo", "axon_value")
LHONfull<-merge(x = LHONaxon, y=LHONden, by = "Combo")
LHONfull<-filter(LHONfull, axon_value!=100 | den_value!=100)
#Run a regression
fit <- lm(formula = axon_value~den_value, data = LHONfull)
summary(fit)
#Plot of the data with regression line 
g<-ggplot(data = LHONfull, mapping = aes(x=axon_value, y=den_value))
g<-g+geom_point(alpha=0.1, size=2)
g<-g+geom_smooth(method='lm', formula = y ~ x)
g<-g+xlab("LHON axonal overlap %")
g<-g+ylab("LHON dendritic overlap %")
g
ggsave(filename = "FigureS4/LHONaxon_LHONden_corr.pdf")

```