library(ggplot2)
library(scales)
library(gridExtra)
library(grid)


##################
### BAR COLORS ###
##################

gg_color_hue <- function(n) {
  hues = seq(15, 375, length = n + 1)
  hcl(h = hues, l = 65, c = 100)[1:n]
}
cols = gg_color_hue(8)

uncorrected = cols[1]
hapog       = rep(cols[2], 6)
hypo        = rep(cols[3], 6)
nextpolish  = rep(cols[4], 6)
ntedit      = rep(cols[5], 6)
pilon       = rep(cols[6], 6)
polca       = rep(cols[7], 6)
racon       = rep(cols[8], 6)


#######################
### DATA FORMATTING ###
#######################

dat = read.csv('metag.all.res')
colnames(dat)[1] = "Tool"
dat$Tool = factor(dat$Tool, levels=c("Uncorrected", "Hapo-G", "HyPo", "NextPolish", "ntEdit", "Pilon", "POLCA", "Racon"))
dat$Round = factor(dat$Round, levels=as.character(unique(dat$Round)))


#####################
### MAPPED PAIRS  ###
#####################

ppairs_top = ggplot(data=dat, aes(x=Tool, y=perfectPairs, group=Round)) +
  geom_col(position=position_dodge2(width = 0.9, preserve = "single"), color="black", fill=c(uncorrected, hapog, hypo, nextpolish, ntedit, pilon, polca, racon)) +
  theme_bw() +
  theme(axis.text = element_text(size = 13), axis.title = element_text(size = 13),legend.text = element_text(size = 18),plot.title = element_text(size = 18)) +
  theme(legend.position = "none")+
  xlab("") +
  ylab("") +
  scale_y_continuous(breaks=c(0, 140000000, 145000000, 150000000), labels=c("0", "140", "145", "150")) +
  coord_cartesian(ylim=c(140000000, 151000000))

ppairs_bottom = ggplot(data=dat, aes(x=Tool, y=perfectPairs, group=Round)) +
  geom_col(position=position_dodge2(width = 0.9, preserve = "single"), color="black", fill=c(uncorrected, hapog, hypo, nextpolish, ntedit, pilon, polca, racon)) +
  theme_bw() +
  theme(axis.text = element_text(size = 13), axis.title = element_text(size = 13),legend.text = element_text(size = 18),plot.title = element_text(size = 18)) +
  theme(legend.position = "none")+
  xlab("") +
  ylab("") +
  scale_y_continuous(breaks=c(0, 25000000, 50000000, 75000000, 100000000, 125000000, 150000000, 175000000), labels=c("0", "25", "50", "75", "100", "125", "150", "175"))

ppairs = grid.arrange(ppairs_top, ppairs_bottom,
                      heights=c(2, 3),
                      ncol=1, nrow=2,
                      left=textGrob("Million of perfectly mapped pairs",
                                    rot=90,
                                    vjust=1,
                                    gp = gpar(fontsize = 18)))


#####################
### SPECIFIC KMERS###
#####################

kmers_top = ggplot(data=dat, aes(x=Tool, y=specific, group=Round)) +
  geom_histogram(stat="identity", position=position_dodge2(width = 0.9, preserve = "single"), color="black", fill=c(uncorrected, hapog, hypo, nextpolish, ntedit, pilon, polca, racon)) +
  theme_bw() +
  theme(axis.text = element_text(size = 13), axis.title = element_text(size = 13),legend.text = element_text(size = 18),plot.title = element_text(size = 18)) +
  theme(legend.position = "none") +
  xlab("") +
  ylab("") +
  scale_y_continuous(breaks=c(0, 43500000, 43600000, 43700000, 43800000), labels=c("0", "43.5M", "43.6M", "43.7M", "43.8M")) +
  coord_cartesian(ylim=c(43500000, 43800000))

kmers_bottom = ggplot(data=dat, aes(x=Tool, y=specific, group=Round)) +
  geom_histogram(stat="identity", position=position_dodge2(width = 0.9, preserve = "single"), color="black", fill=c(uncorrected, hapog, hypo, nextpolish, ntedit, pilon, polca, racon)) +
  theme_bw() +
  theme(axis.text = element_text(size = 13), axis.title = element_text(size = 13),legend.text = element_text(size = 18),plot.title = element_text(size = 18)) +
  theme(legend.position = "none") +
  xlab("") +
  ylab("") +
  scale_y_continuous(breaks=c(0, 10000000, 20000000, 30000000, 40000000, 50000000), labels=c("0", "10M", "20M", "30M", "40M", "50M"))

kmers = grid.arrange(kmers_top, kmers_bottom,
                      heights=c(2, 3),
                      ncol=1, nrow=2,
                      left=textGrob("Number of specific 31-mers",
                                    rot=90,
                                    vjust=1,
                                    gp = gpar(fontsize = 18)))



################
### RUN TIME ###
################

dat2 = subset(dat, dat$Tool != "Uncorrected")

run_time = ggplot(data=dat2, aes(x=Tool, y=elapsed, fill=Tool, group=Round)) +
  geom_histogram(stat="identity", position=position_dodge(), color="black", fill=c(hapog, hypo, nextpolish, ntedit, pilon, polca, racon)) +
  theme_bw() +
  theme(axis.text = element_text(size = 16), axis.title = element_text(size = 16),legend.text = element_text(size = 18),plot.title = element_text(size = 18)) +
  theme(legend.position = "none") +
  xlab("") +
  ylab("Total running time (minutes)")+
  scale_y_continuous(limits=c(0, 500))

ppairs <- arrangeGrob(ppairs, top = textGrob("A", x = unit(0, "npc"),
                                             y   = unit(1, "npc"), just=c("left","top"),
                                             gp=gpar(col="black", fontsize=18, fontfamily="Times Roman")))

kmers <- arrangeGrob(kmers, top = textGrob("B", x = unit(0, "npc"),
                                             y = unit(1, "npc"), just=c("left","top"),
                                             gp=gpar(col="black", fontsize=18, fontfamily="Times Roman")))

run_time <- arrangeGrob(run_time, top = textGrob("C", x = unit(0, "npc"),
                                                 y = unit(1, "npc"), just=c("left","top"),
                                                 gp=gpar(col="black", fontsize=18, fontfamily="Times Roman")))

g <- grid.arrange(ppairs, kmers, run_time, ncol=2, nrow=2)
ggsave(file="Figure_MetaG.png", plot=g, width=16, height=9)