Figure6andS6.Rmd

---
title: "Figure6"
author: "Mike"
---
#Figure6 Plotting 
Code to plot the MBON and DAN heatmaps, for both Figure6A/B and the supplementary information

First load up the required libraries and functions
```{r, echo=FALSE, results="hide"}
library(here)
library(dplyr)
library(xlsx)
library(NMF)                # aheatmap()
library(ggplot2)
readin<-function(file) {
  as.matrix(read.table(file, header=TRUE,row.names=1))
}
here()
file.symlink("/Users/dolanm/Google Drive/SplitGAL4_Paper/Figures/Figure6", here())
```

Read in the MBON and LHON neurotransmitter annotations 
Note: Use black heatmaps so the colours aren't so overwhealming
Note: Read in the NAmaster from dropbox for now. Will use an updated final version for the markdown
Prelinimary code chunk to read in the NT tracks 
```{r, echo=FALSE}
#Load up the revelent datafiles 
LHON_MBON<-readin(file = "Figure6/Output_axonmemb_x_MBON_axon.txt")
MBONnts<-read.xlsx(file = "Figure6/MBON_NT.xlsx",sheetIndex = 1)
NAmaster<-read.xlsx(file="/Users/dolanm/Dropbox/JFRCvisitorProject/Neuroanatomy_Master.xlsx", sheetIndex = 1)
LHnts<-select(NAmaster,LHClusters., Neurotransmitter)
LHnewnames<-select(NAmaster,LHClusters., FinalNames)

#The annotation::cell-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_MBON)[[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))

#Do the same for MBONs 
MBONs<-dimnames(LHON_MBON)[[2]] #Pull in the cell-types in question and get their order 
MBONs<-data.frame(Type=MBONs)
names(MBONnts)<-c("Type", "Neurotransmitter")

MBONtransmitter<-left_join(x = MBONs, y = MBONnts, "Type")
MBONtransmitter<-data.frame(Neurotransmitter=as.character(MBONtransmitter$Neurotransmitter))

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

#Change the names of LH anatomy groups to their cell-types 
oldnames<-lapply(X = dimnames(LHON_MBON), 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_MBON)<-list(newname1,gsub(dimnames(LHON_MBON)[[2]], pattern = "CellType_", replacement = "")) 

#Plot the MBON and LHONs with neurotransmitter tracks 
aheatmap(x = LHON_MBON, filename=paste0("Figure6/LHON_MBON.pdf")
         ,fontsize = 10, cexRow = 3, col="black",annRow = LHONtransmitter, annCol=MBONtransmitter, annColors  = list(transmitter_col), cexCol = 0.65, treeheight = 12, Colv = FALSE, annLegend = FALSE)
```

Load and plot the overlap analysis of LHON axons and DAN dendrites 
```{r, echo=FALSE}
LHON_DAN<-readin(file = "Figure6/Output_axonmemb_x_DAN_den.txt")

dimnames(LHON_DAN)<-list(newname1,gsub(dimnames(LHON_DAN)[[2]], pattern = "CellType_", replacement = "")) 
aheatmap(x = LHON_DAN, filename=paste0("Figure6/LHON_DAN.pdf")
         ,fontsize = 10, cexRow = 3, col="black",annRow = LHONtransmitter, annColors  = list(transmitter_col),  cexCol = 0.65, treeheight = 12, Colv = FALSE, annLegend = FALSE)
```

Calculate the number of LHONs that interact with MB-associated neurons. Use a threshold of 15% as per V2 paper  
```{r, echo=FALSE}
full.mb<-cbind(LHON_DAN, LHON_MBON)
full.mb<-(full.mb>15)
full.mb[full.mb==TRUE]<-1
table(rowSums(full.mb))
```

```{r, echo=FALSE}


```