########## YOU NEED TO SPECIFY THIS PATH #################
# data.dir <- "PATH" <<<<<< ! ! ! 
############## ########## ########## ########## ##########
stopifnot(file.exists(data.dir))

d1 <- "EL_wood_compare_4_isolates.txt"
data1 <- read.table(file=paste0(data.dir, "/", d1), skip=4, header=TRUE, sep="\t", quote="")
head(data1)

summary(data1)

d2 <- "EL_wood_compare_5_cultivars_Vv.txt"
data2 <- read.table(file=paste0(data.dir, "/", d2), skip=4, header=TRUE, sep="\t", quote="")
head(data2)

summary(data2)

dpi <- "EL_qPCR_different_points.txt"
dpi.df<- read.table(file=paste0(data.dir, "/", dpi), header=TRUE, skip=2, sep="\t")
head(dpi.df)

summary(dpi.df)

data.agr <- paste0(data.dir,"/data_AGR.csv")
d.agr <- read.table(file=data.agr, header=TRUE, sep=";")
head(d.agr)

summary(d.agr)

# Function to calculate the mean and the standard deviation and percentage (NA removed) 
# for each group
# data : a data frame
# varname : the name of a column containing the variable to be summarized
# groupnames : vector of column names to be used as grouping variables

data_summary <- function(data, varname, groupnames){
  require(plyr)
  summary_func <- function(x, col){
                  c(mean = mean(x[[col]], na.rm=TRUE),
                    sd = sd(x[[col]], na.rm=TRUE),
                    numb = sum(x[[col]], na.rm=TRUE),
                    perc = round(sum(x[[col]], na.rm=TRUE)
                                 /length(na.omit(x[[col]]))*100, 
                                 digits = 0),
                    count.no.NA = length(na.omit(x[[col]]))
                    ) 
  }
  data_sum <- ddply(data, groupnames, .fun=summary_func, varname)
  #data_sum <- rename(data_sum, c("mean" = varname))
 return(data_sum)
}

# Function to calculate the standard error from dataframe x
  std.error <- function(x, na.rm = T) {
      sqrt(var(x, na.rm = na.rm)/length(x[complete.cases(x)]))
  }

# load data
  df<- data1

# update labelling in the dataframe 
  levels(df$EL)<- c("UNCONTAMINATED","EL","CONTAMINATED")
  df$EL <- factor(df$EL, levels = c("EL","UNCONTAMINATED","CONTAMINATED"))

# barplot 
  p <- ggplot(data=na.omit(df), aes(dist_cm, fill=as.factor(EL))) + geom_bar()
  p <- p + stat_count(aes(label = ..count..), geom="text", color="black", vjust=1.1)
  p <- p + xlab("Distance (cm) from inoculation site")
  p <- p + ylab("Number of observations")
  p <- p + scale_fill_manual(values=c("#66CC99", "grey90","#FF67A4"))
  p <- p + facet_grid(dpi ~ treat)
  p <- p + theme_bw() + theme(legend.position="bottom")
  print(p)  

# load data
  df<- data2

# replace "C" by "conta" and !="C" by "uncontam." in df
  levels(df$EL)<- c("UNCONTAMINATED","EL","CONTAMINATED")
  df$EL <- factor(df$EL, levels = c("EL","UNCONTAMINATED","CONTAMINATED"))
  
# select data from controls
  tmp <- subset(na.omit(df),treat=="T-")
  levels(tmp$treat)[1]<- "CONTROLS"

# plot results 
  p <- ggplot(data=tmp, aes(dist_cm, fill=as.factor(EL))) + geom_bar()
  p <- p + stat_count(aes(label = ..count..), geom="text", color="black", vjust=1.1)
  p <- p + xlab("Distance from inoculation site")+ ylab("Number of observations")
  p <- p + scale_fill_manual(values=c("grey90","#FF67A4"))
  p <- p + facet_grid(dpi ~ var)+ scale_x_discrete(limits=0:2)
  p <- p + theme_bw() + theme(legend.position="bottom")
print(p)  

# remove controls
  tmp <- subset(na.omit(df),treat!="T-")

# plot results 
  p <- ggplot(data=tmp, aes(dist_cm, fill=as.factor(EL))) + geom_bar()
  p <- p + stat_count(aes(label = ..count..), geom="text", color="black", vjust=1.1)
  p <- p + xlab("Distance from inoculation site")+ ylab("Number of observations")
  p <- p + scale_fill_manual(values=c("#66CC99", "grey90","#FF67A4"))
  p <- p + facet_grid(dpi ~ var)+ scale_x_discrete(limits=0:2)
  p <- p + theme_bw() + theme(legend.position="bottom")
print(p) 

# Replace contaminated sample by "NA" in the dataframe
  data2$EL[data2$EL == "C"] <- NA                        # replace "C" by "NA"
  data2$EL <- as.numeric(as.character(data2$EL))         # as numeric (0/1/NA)
  data2$dpi <- as.factor(as.character(data2$dpi))        # as numeric (15/30/45/60)
  data2$dist_cm <- as.factor(as.character(data2$dist_cm))# as numeric (0/1/2)

# calculate the mean and the standard deviation and percentage (NA removed)
  data2_sum <- data_summary(data= data2, varname= "EL",
                           groupnames=c("dpi","treat","var","dist_cm"))
# update column names
  colnames(data2_sum)[5]<- "EL.mean"
  colnames(data2_sum)[6]<- "EL.mean.sd"
  colnames(data2_sum)[8]<- "EL.perc"

# select only inoculations
  d <- subset(data2_sum, data2_sum$treat != "T-")
  d <- subset(d, d$dist_cm != "0")
  head(d)

### FUNCTION to plot colonization of wood per variety

  ggplot.wood.coloniz.per.dist <- function(d, distance){
  
        dtemp <-subset(d, d$dist_cm==distance)
  
  # ggplot
  p <- ggplot(dtemp, aes(x=dpi, y=EL.perc, group=var)) 
  p <- p + geom_line(aes(linetype=var, color=var, size=var))
  p <- p + geom_point(aes(shape=var), size=3)
  p <- p + scale_linetype_manual(values=c("solid","dashed","dashed","solid","solid"))
  p <- p + scale_color_manual(values= c("red","violet","green","brown","blue"))
  p <- p + scale_size_manual(values=c(1,1,1,1,1))
  p <- p + scale_shape_manual(values=c(17,3,15,19,19))
  p <- p + labs(title = paste0("at ",distance, " cm from IP"))
  p <- p + ylim(c(0, 100))
  p <- p + theme_bw() +  theme(legend.position="bottom")
  print(p)
}

p1 <- ggplot.wood.coloniz.per.dist(d, distance=1)

p2 <- ggplot.wood.coloniz.per.dist(d, distance=2)

# Function to calculate the standard error from dataframe x
  std.error <- function(x, na.rm = T) {
      sqrt(var(x, na.rm = na.rm)/length(x[complete.cases(x)]))
  }

# calculate means and std.error 
  dpi.mean <- ddply(dpi.df, .(treat, PCR, dist_cm, dpi), summarise,
                    "nb_copies.mean"= mean(nb_copies, 2, na.rm = TRUE),
                    "nb_copies.std" = std.error(nb_copies, na.rm = TRUE),
                    count.no.NA = length(na.omit(nb_copies)))
  str(dpi.mean)

  head(dpi.mean)

# remove actin PCR
  tmp <- subset(dpi.mean, PCR!="PAL" & PCR!="PCH" &
                treat!="H2O" & treat!="PAL" & treat!="PCH")
  levels(tmp$treat)[5]<- "CONTROLS"
  levels(tmp$treat)[1]<- "EL inoc"
  levels(tmp$PCR)[1] <- "Actin V.vinifera"
  levels(tmp$PCR)[2] <- "Beta-tubulin E.lata"
  levels(tmp$dist_cm) <- c("-1 cm below IP", "+1 cm above IP")

# colors
  my_col=c("grey","pink")
    
# Define the top and bottom of the errorbars
  limits <- aes(ymax= nb_copies.mean + nb_copies.std, 
                ymin= nb_copies.mean - nb_copies.std)

# barplot and std.error
  p <- ggplot(tmp, aes(as.factor(dpi), nb_copies.mean, fill=PCR)) 
  #p <- p + ylim(0,40000)
  p <- p + geom_bar(stat="identity", position="dodge", color="black")
  p <- p + geom_errorbar(limits, position=position_dodge(.9),width=.2)
  p <- p + geom_text(aes(label=paste0(round(nb_copies.mean,0),"\n(n=",count.no.NA,")")),
                     position=position_dodge(width=0.9), vjust=-.1)
  p <- p + scale_fill_manual(values = my_col)
  p <- p + labs(title= "", x= "Number of days after inoculation", 
                y= "Gene copy number (mean), by qPCR") 
  p <- p + facet_grid(PCR ~ dist_cm*treat, scales = "free_y") 
  p <- p + theme_bw() + theme(legend.position="top")
print(p)

# format data before ratio
  d <- dpi.df[,c("sample","treat","PCR","dist_cm","dpi","nb_copies")]
  d$dpi <- factor(d$dpi)
  
# aggregate
  d2 <- cast(d, sample+treat+dist_cm+dpi ~ PCR, median)

# calculate ratio    
  d2$ratio <- d2$EL / d2$actin

# calculate RATIO means and std.error 
  d2.mean <- ddply(d2, .(treat, dist_cm, dpi), summarise,
                   "ratio.mean"= mean(ratio, 2, na.rm = TRUE),
                   "ratio.std" = std.error(ratio, na.rm = TRUE),
                   "count" = length(na.omit(ratio)))
  
# Define the top and bottom of the errorbars
  limits <- aes(ymax= ratio.mean + ratio.std, 
                ymin= ratio.mean - ratio.std)

# barplot
  p <- ggplot(na.omit(d2.mean), aes(as.factor(dpi), ratio.mean, fill=treat)) 
  p <- p + geom_bar(stat="identity", position="dodge", color="black")
  p <- p + ylim(0,0.75)
  p <- p + geom_errorbar(limits, position=position_dodge(.9),width=.2)
  p <- p + geom_text(aes(label=paste0(round(ratio.mean,2),"\n(n=",count,")")),
                     position=position_dodge(width=0.9), vjust=-.1)
  p <- p + scale_fill_manual(values = rep("lightblue",3))
  p <- p + labs(title= "", x= "Number of days after inoculation",
                y= "Ratio b-tubulin/actin")
  p <- p + facet_grid(. ~ dist_cm*treat, scales = "free_y") 
  p <- p + theme_bw() + theme(legend.position="none")
print(p)

# select data
  d <- d.agr

# Set marging
  par(mar = c(7, 4, 5, 1))

# Boxplot tubulin vs treatment
  boxplot(d$btub_Elata ~ d$Treatment,
          main=paste0("Comparison of E.lata isolates \nfor wood colonization"),
          ylab="Genome copy number", xlab="E. lata isolates",
          varwidth= TRUE, notch= FALSE, las= 2, col= c("brown", "green"))

# Add points
  points(d$btub_Elata ~ d$Treatment, pch= 19)