getwd()
setwd("~/git/LSHTM_analysis/mcsm_analysis/pyrazinamide/scripts/plotting")
getwd()

############################################################
# 1: Installing and loading required packages and functions
############################################################

#source("../Header_TT.R")
#source("../barplot_colour_function.R")
#library(tidyverse)

###########################
#2: Read file: normalised file, output of step 4 mcsm pipeline
###########################
#my_df <- read.csv("../../Data/mcsm_complex1_normalised.csv"
#                  , row.names = 1
#                  , stringsAsFactors = F
#                  , header = T)

# call script combining_df
source("../combining_two_df.R")

#---------------------- PAY ATTENTION
# the above changes the working dir
#  from Plotting to Scripts"
#---------------------- PAY ATTENTION

#==========================
# This will return:

# df with NA for pyrazinamide:
#merged_df2 
#merged_df2_comp 

# df without NA for pyrazinamide:
#merged_df3 
#merged_df3_comp 
#==========================
###########################
# Data to choose: 
# We will be using the small dfs
# to generate the coloured axis
###########################

# uncomment as necessary
#!!!!!!!!!!!!!!!!!!!!!!!
# REASSIGNMENT
my_df  = merged_df3
#my_df  = merged_df3_comp
#!!!!!!!!!!!!!!!!!!!!!!!

# delete variables not required
rm(merged_df2, merged_df2_comp, merged_df3, merged_df3_comp)

str(my_df)
my_df$Position
c1 = my_df[my_df$Mutationinformation == "L4S",]

# order my_df by Position
my_df_o = my_df[order(my_df$Position),]
head(my_df_o$Position); tail(my_df_o$Position)

c2 = my_df_o[my_df_o$Mutationinformation == "L4S",]

# sanity check
if (sum(table(c1 == c2)) == ncol(my_df)){
  print ("Sanity check passsd")
}else{
  print ("Error!: Please debug your code")
}

rm(my_df, c1, c2)

# create a new df with unique position numbers and cols
Position = unique(my_df_o$Position) #130
Position_cols = as.data.frame(Position)

head(Position_cols) ; tail(Position_cols)

# specify active site residues and bg colour
Position = c(49, 51, 57, 71
             , 8, 96, 138
             , 13, 68
             , 103, 137
             , 133, 134) #13

lab_bg = rep(c("purple" 
               , "yellow"
               , "cornflowerblue"
               , "blue"
               , "green"), times = c(4, 3, 2, 2, 2)
)

# second bg colour for active site residues
#lab_bg2 = rep(c("white" 
#                , "green" , "white", "green"
#                , "white"
#                , "white"
#                , "white"), times = c(4
#                                      , 1, 1, 1
#                                      , 2
#                                      , 2
#                                      , 2)
#)

#%%%%%%%%%
# revised: leave the second box coloured as the first one incase there is no second colour
#%%%%%%%%%
lab_bg2 = rep(c("purple" 
                , "green", "yellow", "green"
                , "cornflowerblue"
                , "blue"
                , "green"), times = c(4
                                      , 1, 1, 1
                                      , 2
                                      , 2
                                      , 2))

# fg colour for labels for active site residues
lab_fg = rep(c("white" 
               , "black"
               , "black"
               , "white"
               , "black"), times = c(4, 3, 2, 2, 2))

#%%%%%%%%%
# revised: make the purple ones black
# fg colour for labels for active site residues
#%%%%%%%%%
#lab_fg = rep(c("black" 
#               , "black"
#               , "black"
#               , "white"
#               , "black"), times = c(4, 3, 2, 2, 2))               
               
# combined df with active sites, bg and fg colours
aa_cols_ref = data.frame(Position
                         , lab_bg
                         , lab_bg2
                         , lab_fg
                         , stringsAsFactors = F) #13, 4

str(Position_cols); class(Position_cols)
str(aa_cols_ref); class(aa_cols_ref)

# since Position is int and numeric in the two dfs resp, 
# converting numeric to int for consistency
aa_cols_ref$Position = as.integer(aa_cols_ref$Position)
class(aa_cols_ref$Position)

#===========
# Merge 1: merging Positions df (Position_cols) and
# active site cols (aa_cols_ref)
# linking column: "Position"
# This is so you can have colours defined for all 130 positions
#===========
head(Position_cols$Position); head(aa_cols_ref$Position)

mut_pos_cols = merge(Position_cols, aa_cols_ref
            , by = "Position"
            , all.x = TRUE) 

head(mut_pos_cols)
# replace NA's 
# :column "lab_bg" with "white"
# : column "lab_fg" with "black"
mut_pos_cols$lab_bg[is.na(mut_pos_cols$lab_bg)] <- "white"
mut_pos_cols$lab_bg2[is.na(mut_pos_cols$lab_bg2)] <- "white"
mut_pos_cols$lab_fg[is.na(mut_pos_cols$lab_fg)] <- "black"
head(mut_pos_cols)

#===========
# Merge 2: Merge mut_pos_cols with mcsm df
# Now combined the 130 positions with aa colours with 
# the mcsm_data
#===========
# dfs to merge
df0 = my_df_o
df1 = mut_pos_cols

# check the column on which merge will be performed
head(df0$Position); tail(df0$Position)
head(df1$Position); tail(df1$Position)

# should now have 3 extra columns
my_df = merge(df0, df1
              , by = "Position"
              , all.x = TRUE)

# sanity check
my_df[my_df$Position == "49",]
my_df[my_df$Position == "13",]

my_df$Position

# clear variables
rm(aa_cols_ref
   , df0
   , df1
   , my_df_o
   , Position_cols
   , lab_bg
   , lab_bg2
   , lab_fg
   , Position
   )