moved not required plots to scratch
This commit is contained in:
parent
923cad81b5
commit
b63bbd6f15
9 changed files with 2 additions and 1660 deletions
|
|
@ -188,6 +188,6 @@ OutPlot_lig_pos_count = g + geom_bar(aes (alpha = 0.5)
|
|||
print(OutPlot_lig_pos_count)
|
||||
dev.off()
|
||||
########################################################################
|
||||
# end of lig barplots
|
||||
# end of LIG barplots
|
||||
########################################################################
|
||||
|
||||
|
|
|
|||
|
|
@ -186,5 +186,5 @@ OutPlot_pos_count = g + geom_bar(aes (alpha = 0.5)
|
|||
print(OutPlot_pos_count)
|
||||
dev.off()
|
||||
########################################################################
|
||||
# end of Ligand barplots
|
||||
# end of PS barplots
|
||||
########################################################################
|
||||
|
|
|
|||
|
|
@ -1,166 +0,0 @@
|
|||
#!/usr/bin/env Rscript
|
||||
#########################################################
|
||||
# TASK: Corr plots for PS and Lig
|
||||
|
||||
# Output: 1 svg
|
||||
|
||||
#=======================================================================
|
||||
# working dir and loading libraries
|
||||
getwd()
|
||||
setwd("~/git/LSHTM_analysis/scripts/plotting/")
|
||||
getwd()
|
||||
|
||||
#source("combining_dfs_plotting.R")
|
||||
source("my_pairs_panel.R") # with lower panel turned off
|
||||
source("corr_data.R")
|
||||
|
||||
#=======
|
||||
# output
|
||||
#=======
|
||||
# PS
|
||||
corr_ps_all_df2 = "corr_PS_ALL_df2.svg"
|
||||
plot_corr_ps_all_df2 = paste0(plotdir,"/", corr_ps_all_df2)
|
||||
|
||||
corr_ps_all_df3 = "corr_PS_ALL_df3.svg"
|
||||
plot_corr_ps_all_df3 = paste0(plotdir,"/", corr_ps_all_df3)
|
||||
|
||||
# LIG
|
||||
corr_lig_all_df2 = "corr_LIG_ALL_df2.svg"
|
||||
plot_corr_lig_all_df2 = paste0(plotdir,"/", corr_lig_all_df2)
|
||||
|
||||
corr_lig_all_df3 = "corr_LIG_ALL_df3.svg"
|
||||
plot_corr_lig_all_df3 = paste0(plotdir,"/", corr_lig_all_df3)
|
||||
|
||||
####################################################################
|
||||
# end of loading libraries and functions
|
||||
####################################################################
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
# Data for plots
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
|
||||
# PS
|
||||
corr_ps_df2 = corr_ps_df2[-1]
|
||||
corr_ps_df3 = corr_ps_df3[-1]
|
||||
|
||||
# Lig
|
||||
corr_lig_df2 = corr_lig_df2[-1]
|
||||
corr_lig_df3 = corr_lig_df3[-1]
|
||||
|
||||
#---------------------------------------
|
||||
# generate corr PS plot 1: merged_df2
|
||||
#---------------------------------------
|
||||
cat("Corr plot PS DUET with coloured dots:", plot_corr_ps_all_df2)
|
||||
svg(plot_corr_ps_all_df2, width = 30, height = 30)
|
||||
|
||||
OutPlot_ps_df2 = pairs.panels(corr_ps_df2[1:(length(corr_ps_df2)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation ellipses
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_ps_df2$duet_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
#, pch = 19
|
||||
, jitter = T
|
||||
, alpha = 1
|
||||
#, points(pch = 19, col = c("#f8766d", "#00bfc4"))
|
||||
, cex = 1.8
|
||||
, cex.axis = 2
|
||||
, cex.labels = 2
|
||||
, cex.cor = 1
|
||||
, smooth = F)
|
||||
|
||||
print(OutPlot_ps_df2)
|
||||
dev.off()
|
||||
|
||||
#----------------------------------------------
|
||||
# generate corr PS plot 2: merged_df3
|
||||
#----------------------------------------------
|
||||
cat("Corr plot PS DUET with coloured dots:", plot_corr_ps_all_df3)
|
||||
svg(plot_corr_ps_all_df3, width = 30, height = 30)
|
||||
|
||||
OutPlot_ps_df3 = pairs.panels(corr_ps_df3[1:(length(corr_ps_df3)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation ellipses
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_ps_df3$duet_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
, cex = 2
|
||||
, cex.axis = 1.6
|
||||
, cex.labels = 2
|
||||
, cex.cor = 1
|
||||
, smooth = F
|
||||
|
||||
)
|
||||
|
||||
print(OutPlot_ps_df3)
|
||||
dev.off()
|
||||
|
||||
################################################################################################
|
||||
|
||||
|
||||
#---------------------------------------
|
||||
# generate corr lig plot 1: merged_df2
|
||||
#---------------------------------------
|
||||
cat("Corr plot lig DUET with coloured dots:", plot_corr_lig_all_df2)
|
||||
svg(plot_corr_lig_all_df2, width = 30, height = 30)
|
||||
|
||||
OutPlot_lig_df2 = pairs.panels(corr_lig_df2[1:(length(corr_lig_df2)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation elliliges
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_lig_df2$ligand_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
#, pch = 19
|
||||
, jitter = T
|
||||
, alpha = 1
|
||||
#, points(pch = 19, col = c("#f8766d", "#00bfc4"))
|
||||
, cex = 1.8
|
||||
, cex.axis = 2
|
||||
, cex.labels = 2
|
||||
, cex.cor = 1
|
||||
, smooth = F)
|
||||
|
||||
print(OutPlot_lig_df2)
|
||||
dev.off()
|
||||
|
||||
#----------------------------------------------
|
||||
# generate corr lig plot 2: merged_df3
|
||||
#----------------------------------------------
|
||||
cat("Corr plot lig DUET with coloured dots:", plot_corr_lig_all_df3)
|
||||
svg(plot_corr_lig_all_df3, width = 30, height = 30)
|
||||
|
||||
OutPlot_lig_df3 = pairs.panels(corr_lig_df3[1:(length(corr_lig_df3)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation elliliges
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_lig_df3$ligand_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
, cex = 2
|
||||
, cex.axis = 1.6
|
||||
, cex.labels = 2
|
||||
, cex.cor = 1
|
||||
, smooth = F
|
||||
|
||||
)
|
||||
|
||||
print(OutPlot_lig_df3)
|
||||
dev.off()
|
||||
|
|
@ -1,176 +0,0 @@
|
|||
#!/usr/bin/env Rscript
|
||||
#########################################################
|
||||
# TASK: Corr plots for PS and Lig
|
||||
|
||||
# Output: 1 svg
|
||||
|
||||
#=======================================================================
|
||||
# working dir and loading libraries
|
||||
getwd()
|
||||
setwd("~/git/LSHTM_analysis/scripts/plotting/")
|
||||
getwd()
|
||||
|
||||
#source("combining_dfs_plotting.R")
|
||||
source("my_pairs_panel.R") # with lower panel turned off
|
||||
source("corr_data.R")
|
||||
|
||||
#=======
|
||||
# output
|
||||
#=======
|
||||
# PS
|
||||
corrplot_ps_df2 = "corr_PS_df2.svg"
|
||||
plot_corr_ps_df2 = paste0(plotdir,"/", corrplot_ps_df2)
|
||||
|
||||
corrplot_ps_df3 = "corr_PS_df3.svg"
|
||||
plot_corr_ps_df3 = paste0(plotdir,"/", corrplot_ps_df3)
|
||||
|
||||
# LIG
|
||||
corrplot_lig_df2 = "corr_LIG_df2.svg"
|
||||
plot_corr_lig_df2 = paste0(plotdir,"/", corrplot_lig_df2)
|
||||
|
||||
corrplot_lig_df3 = "corr_LIG_df3.svg"
|
||||
plot_corr_lig_df3 = paste0(plotdir,"/", corrplot_lig_df3)
|
||||
|
||||
####################################################################
|
||||
# end of loading libraries and functions
|
||||
####################################################################
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
# Data for plots
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
|
||||
cols_to_drop = c("ASA", "AF_kin")
|
||||
|
||||
|
||||
# PS
|
||||
corr_ps_df2 = corr_ps_df2[!colnames(corr_ps_df2)%in%cols_to_drop]
|
||||
corr_ps_df2 = corr_ps_df2[-1]
|
||||
|
||||
corr_ps_df3 = corr_ps_df3[!colnames(corr_ps_df3)%in%cols_to_drop]
|
||||
corr_ps_df3 = corr_ps_df3[-1]
|
||||
|
||||
|
||||
# Lig
|
||||
corr_lig_df2 = corr_lig_df2[!colnames(corr_lig_df2)%in%cols_to_drop]
|
||||
corr_lig_df2 = corr_lig_df2[-1]
|
||||
|
||||
corr_lig_df3 = corr_lig_df3[!colnames(corr_lig_df3)%in%cols_to_drop]
|
||||
corr_lig_df3 = corr_lig_df3[-1]
|
||||
|
||||
#---------------------------------------
|
||||
# generate corr PS plot 1: merged_df2
|
||||
#---------------------------------------
|
||||
cat("Corr plot PS DUET with coloured dots:", plot_corr_ps_df2)
|
||||
svg(plot_corr_ps_df2, width = 30, height = 25)
|
||||
|
||||
OutPlot_ps_df2 = pairs.panels(corr_ps_df2[1:(length(corr_ps_df2)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation ellipses
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_ps_df2$duet_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
#, pch = 19
|
||||
, jitter = T
|
||||
, alpha = 1
|
||||
#, points(pch = 19, col = c("#f8766d", "#00bfc4"))
|
||||
, cex = 1.8
|
||||
, cex.axis = 2
|
||||
, cex.labels = 3.8
|
||||
, cex.cor = 1
|
||||
, smooth = F)
|
||||
|
||||
print(OutPlot_ps_df2)
|
||||
dev.off()
|
||||
|
||||
#----------------------------------------------
|
||||
# generate corr PS plot 2: merged_df3
|
||||
#----------------------------------------------
|
||||
cat("Corr plot PS DUET with coloured dots:", plot_corr_ps_df3)
|
||||
svg(plot_corr_ps_df3, width = 30, height = 25)
|
||||
|
||||
OutPlot_ps_df3 = pairs.panels(corr_ps_df3[1:(length(corr_ps_df3)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation ellipses
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_ps_df3$duet_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
, cex = 3
|
||||
, cex.axis = 1.6
|
||||
, cex.labels = 3.8
|
||||
, cex.cor = 1
|
||||
, smooth = F
|
||||
|
||||
)
|
||||
|
||||
print(OutPlot_ps_df3)
|
||||
dev.off()
|
||||
|
||||
################################################################################################
|
||||
|
||||
#---------------------------------------
|
||||
# generate corr lig plot 1: merged_df2
|
||||
#---------------------------------------
|
||||
cat("Corr plot lig DUET with coloured dots:", plot_corr_lig_df2)
|
||||
svg(plot_corr_lig_df2, width = 30, height = 25)
|
||||
|
||||
OutPlot_lig_df2 = pairs.panels(corr_lig_df2[1:(length(corr_lig_df2)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation elliliges
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_lig_df2$ligand_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
#, pch = 19
|
||||
, jitter = T
|
||||
, alpha = 1
|
||||
#, points(pch = 19, col = c("#f8766d", "#00bfc4"))
|
||||
, cex = 1.8
|
||||
, cex.axis = 2
|
||||
, cex.labels = 3.8
|
||||
, cex.cor = 1
|
||||
, smooth = F)
|
||||
|
||||
print(OutPlot_lig_df2)
|
||||
dev.off()
|
||||
|
||||
#----------------------------------------------
|
||||
# generate corr lig plot 2: merged_df3
|
||||
#----------------------------------------------
|
||||
cat("Corr plot lig DUET with coloured dots:", plot_corr_lig_df3)
|
||||
svg(plot_corr_lig_df3, width = 30, height = 25)
|
||||
|
||||
OutPlot_lig_df3 = pairs.panels(corr_lig_df3[1:(length(corr_lig_df3)-2)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = T # show density plots
|
||||
, ellipses = F # show correlation elliliges
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(corr_lig_df3$ligand_outcome))] # can't use colour as duet and foldx are opposite
|
||||
, pch = 21 # for bg
|
||||
, cex = 3
|
||||
, cex.axis = 1.6
|
||||
, cex.labels = 3.8
|
||||
, cex.cor = 1
|
||||
, smooth = F
|
||||
|
||||
)
|
||||
|
||||
print(OutPlot_lig_df3)
|
||||
dev.off()
|
||||
|
||||
|
|
@ -1,191 +0,0 @@
|
|||
#!/usr/bin/env Rscript
|
||||
#########################################################
|
||||
# TASK: Corr plots for PS and Lig
|
||||
|
||||
# Output: 1 svg
|
||||
|
||||
#=======================================================================
|
||||
# working dir and loading libraries
|
||||
getwd()
|
||||
setwd("~/git/LSHTM_analysis/scripts/plotting/")
|
||||
getwd()
|
||||
|
||||
|
||||
source("Header_TT.R")
|
||||
require(cowplot)
|
||||
source("combining_dfs_plotting.R") # FIXME: add extra from other plots here
|
||||
|
||||
# should return the following dfs, directories and variables
|
||||
|
||||
|
||||
#=======
|
||||
# output
|
||||
#=======
|
||||
# can't combine by cowplot because not ggplots
|
||||
#corr_plot_combined = "corr_combined.svg"
|
||||
#plot_corr_plot_combined = paste0(plotdir,"/", corr_plot_combined)
|
||||
|
||||
# PS foldx
|
||||
corr_foldx = "corr_adjusted_foldx.svg"
|
||||
plot_corr_foldx = paste0(plotdir,"/", corr_foldx)
|
||||
|
||||
####################################################################
|
||||
# end of loading libraries and functions #
|
||||
########################################################################
|
||||
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
df_ps = merged_df3
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
|
||||
rm( merged_df2, merged_df2_comp, merged_df2_lig
|
||||
, merged_df2_comp_lig
|
||||
, merged_df3_comp, merged_df3_comp_lig
|
||||
, my_df_u, my_df_u_lig)
|
||||
|
||||
########################################################################
|
||||
# end of data extraction and cleaning for plots #
|
||||
########################################################################
|
||||
|
||||
#===========================
|
||||
# Data for Correlation plots:foldx
|
||||
#===========================
|
||||
#============================
|
||||
# adding foldx scaled values
|
||||
# scale data b/w -1 and 1
|
||||
#============================
|
||||
n = which(colnames(df_ps) == "ddg"); n
|
||||
|
||||
my_min = min(df_ps[,n]); my_min
|
||||
my_max = max(df_ps[,n]); my_max
|
||||
|
||||
df_ps$foldx_scaled = ifelse(df_ps[,n] < 0
|
||||
, df_ps[,n]/abs(my_min)
|
||||
, df_ps[,n]/my_max)
|
||||
# sanity check
|
||||
my_min = min(df_ps$foldx_scaled); my_min
|
||||
my_max = max(df_ps$foldx_scaled); my_max
|
||||
|
||||
if (my_min == -1 && my_max == 1){
|
||||
cat("PASS: foldx ddg successfully scaled b/w -1 and 1"
|
||||
, "\nProceeding with assigning foldx outcome category")
|
||||
}else{
|
||||
cat("FAIL: could not scale foldx ddg values"
|
||||
, "Aborting!")
|
||||
}
|
||||
|
||||
|
||||
#================================
|
||||
# adding foldx outcome category
|
||||
# ddg<0 = "Stabilising" (-ve)
|
||||
#=================================
|
||||
|
||||
c1 = table(df_ps$ddg < 0)
|
||||
df_ps$foldx_outcome = ifelse(df_ps$ddg < 0, "Stabilising", "Destabilising")
|
||||
c2 = table(df_ps$ddg < 0)
|
||||
|
||||
if ( all(c1 == c2) ){
|
||||
cat("PASS: foldx outcome successfully created")
|
||||
}else{
|
||||
cat("FAIL: foldx outcome could not be created. Aborting!")
|
||||
exit()
|
||||
}
|
||||
|
||||
table(df_ps$foldx_outcome)
|
||||
|
||||
#======================
|
||||
# adding log cols
|
||||
#======================
|
||||
df_ps$log10_or_mychisq = log10(df_ps$or_mychisq)
|
||||
df_ps$neglog_pval_fisher = -log10(df_ps$pval_fisher)
|
||||
|
||||
df_ps$log10_or_kin = log10(df_ps$or_kin)
|
||||
df_ps$neglog_pwald_kin = -log10(df_ps$pwald_kin)
|
||||
|
||||
# subset data to generate pairwise correlations
|
||||
cols_to_select_foldx = c("foldx_scaled"
|
||||
|
||||
, "duet_scaled"
|
||||
|
||||
, "log10_or_mychisq"
|
||||
, "neglog_pval_fisher"
|
||||
|
||||
, "log10_or_kin"
|
||||
, "neglog_pwald_kin"
|
||||
|
||||
, "af"
|
||||
|
||||
, "foldx_outcome"
|
||||
, drug)
|
||||
|
||||
corr_data_foldx = df_ps[, cols_to_select_foldx]
|
||||
|
||||
dim(corr_data_foldx)
|
||||
|
||||
#p_italic = substitute(paste("-Log(", italic('P'), ")"));p_italic
|
||||
#p_adjusted_italic = substitute(paste("-Log(", italic('P adjusted'), ")"));p_adjusted_italic
|
||||
|
||||
# assign nice colnames (for display)
|
||||
my_corr_colnames_foldx = c("Foldx"
|
||||
|
||||
,"DUET"
|
||||
|
||||
, "Log(OR)"
|
||||
, "-Log(P)"
|
||||
|
||||
, "Log(OR adjusted)"
|
||||
, "-Log(P wald)"
|
||||
|
||||
, "AF"
|
||||
|
||||
, "foldx_outcome"
|
||||
, drug)
|
||||
|
||||
length(my_corr_colnames_foldx)
|
||||
|
||||
colnames(corr_data_foldx)
|
||||
colnames(corr_data_foldx) <- my_corr_colnames_foldx
|
||||
colnames(corr_data_foldx)
|
||||
|
||||
#-----------------
|
||||
# generate corr foldx plot
|
||||
#-----------------
|
||||
start = 1
|
||||
end = which(colnames(corr_data_foldx) == drug); end # should be the last column
|
||||
offset = 1
|
||||
|
||||
my_corr_foldx = corr_data_foldx[start:(end-offset)]
|
||||
head(my_corr_foldx)
|
||||
|
||||
#my_cols = c("#f8766d", "#00bfc4")
|
||||
# deep blue :#007d85
|
||||
# deep red: #ae301e
|
||||
|
||||
cat("Corr plot foldx:", plot_corr_foldx)
|
||||
svg(plot_corr_foldx, width = 15, height = 15)
|
||||
|
||||
OutPlot_foldx= pairs.panels(my_corr_foldx[1:(length(my_corr_foldx)-1)]
|
||||
, method = "spearman" # correlation method
|
||||
, hist.col = "grey" ##00AFBB
|
||||
, density = TRUE # show density plots
|
||||
, ellipses = F # show correlation ellipses
|
||||
, stars = T
|
||||
, rug = F
|
||||
, breaks = "Sturges"
|
||||
, show.points = T
|
||||
, bg = c("#f8766d", "#00bfc4")[unclass(factor(my_corr_foldx$foldx_outcome))]
|
||||
, pch = 21
|
||||
, jitter = T
|
||||
#, alpha = .05
|
||||
#, points(pch = 19, col = c("#f8766d", "#00bfc4"))
|
||||
, cex = 3
|
||||
, cex.axis = 2.5
|
||||
, cex.labels = 2.1
|
||||
, cex.cor = 1
|
||||
, smooth = F
|
||||
)
|
||||
|
||||
print(OutPlot_foldx)
|
||||
dev.off()
|
||||
|
||||
|
||||
|
||||
|
|
@ -1,289 +0,0 @@
|
|||
|
||||
|
||||
#########################################################
|
||||
# 1: Installing and loading required packages
|
||||
#########################################################
|
||||
|
||||
#source("../Header_TT.R")
|
||||
install.packages("qqman")
|
||||
library(qqman)
|
||||
|
||||
source("combining_dfs_plotting.R")
|
||||
#mcsm_data: raw file, 225, 15
|
||||
#merged_df2 = 2201, 35
|
||||
#merged_df3 = 205, 35 ("Can't trust non-numerical params')
|
||||
|
||||
#===============================================
|
||||
# PLOTS: DUET vs GWAS: non-numerical
|
||||
# lineage, country_code, etc
|
||||
# merged_df2: 1592, 35
|
||||
#===============================================
|
||||
|
||||
#########################
|
||||
# Data for plot
|
||||
#########################
|
||||
df = merged_df2
|
||||
#df = merged_df2_comp
|
||||
|
||||
|
||||
#========================
|
||||
# Plot 1a: Lineage barplot
|
||||
# x = lineage y = No of samples
|
||||
# col = Lineage
|
||||
# fill = lineage
|
||||
#========================
|
||||
table(df$lineage)
|
||||
|
||||
# subset only lineages1-4
|
||||
sel_lineages = c("lineage1"
|
||||
, "lineage2"
|
||||
, "lineage3"
|
||||
, "lineage4"
|
||||
#, "lineage5"
|
||||
#, "lineage6"
|
||||
#, "lineage7"
|
||||
)
|
||||
|
||||
# uncomment as necessary
|
||||
df_lin = subset(df, subset = lineage %in% sel_lineages )
|
||||
table(df_lin$lineage)
|
||||
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
# REASSIGNMENT
|
||||
df <- df_lin
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
# REASSIGNMENT
|
||||
df2 = df
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
df2 = df2%>%
|
||||
add_count(country_code)
|
||||
|
||||
str(df2$country_code); str(df2$n)
|
||||
|
||||
n = which(colnames(df2) == "n")
|
||||
colnames(df2)[n] = "count_country"
|
||||
|
||||
table(df2$count_country>100 & df$country_code!= "")
|
||||
df3 = subset(df2, df2$count_country>100 & df2$country_code != "")
|
||||
|
||||
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
# REASSIGNMENT
|
||||
df = df3
|
||||
#%%%%%%%%%%%%%%%%%%%%%%%%
|
||||
|
||||
sample = sum(table(unique(df$id))); sample
|
||||
table(df$country_code)
|
||||
tab = sum(table(df$country_code)); tab
|
||||
|
||||
|
||||
View(table(df$country_code))
|
||||
View(t1)
|
||||
|
||||
############## begin plot
|
||||
g = ggplot(df, aes(x = lineage))
|
||||
g + geom_bar(aes(fill = lineage)) +
|
||||
theme( axis.text.x = element_text(size = 13
|
||||
, angle = 90
|
||||
, hjust = 1
|
||||
, vjust = 0.4)
|
||||
, axis.text.y = element_text(size = 15
|
||||
, angle = 0
|
||||
, hjust = 1
|
||||
, vjust = 0)
|
||||
, axis.title.x = element_text(size = 15)
|
||||
, axis.title.y = element_text(size = 15) ) +
|
||||
labs(title = "Lineage"
|
||||
, x = "Lineage"
|
||||
, y = "No of samples")
|
||||
|
||||
|
||||
#========================
|
||||
# Plot 2: DUET, lineage, country_code and or_mychisq
|
||||
# x = lineage y = DUET
|
||||
# col = Lineage
|
||||
# fill = country_code
|
||||
#========================
|
||||
### begin plot
|
||||
g = ggplot(df, aes(x = country_code
|
||||
, y = duet_scaled))
|
||||
g + geom_point(aes(col = lineage
|
||||
, size = or_mychisq)) +
|
||||
theme(axis.text.x = element_text(size = 13
|
||||
, angle = 90
|
||||
, hjust = 1
|
||||
, vjust = 0.4)
|
||||
, axis.text.y = element_text(size = 15
|
||||
, angle = 0
|
||||
, hjust = 1
|
||||
, vjust = 0)
|
||||
, axis.title.x = element_text(size = 15)
|
||||
, axis.title.y = element_text(size = 15) ) +
|
||||
labs(title = "DUET, country_code, lineage, or_mychisq"
|
||||
, x = "Lineage"
|
||||
, y = "DUET (PS)")
|
||||
|
||||
|
||||
#############
|
||||
#========================
|
||||
# Plot 3: DUET, lineage, or_mychisq
|
||||
# x = lineage y = DUET
|
||||
# col = Lineage
|
||||
# fill = country_code
|
||||
#========================
|
||||
|
||||
### begin plot
|
||||
table(df$lineage)
|
||||
|
||||
g = ggplot(df_lin, aes(x = lineage
|
||||
, y = duet_scaled))
|
||||
g + geom_point(aes(col = lineage
|
||||
, size = or_mychisq)) +
|
||||
theme(axis.text.x = element_text(size = 13
|
||||
, angle = 90
|
||||
, hjust = 1
|
||||
, vjust = 0.4)
|
||||
, axis.text.y = element_text(size = 15
|
||||
, angle = 0
|
||||
, hjust = 1
|
||||
, vjust = 0)
|
||||
, axis.title.x = element_text(size = 15)
|
||||
, axis.title.y = element_text(size = 15) ) +
|
||||
labs(title = "DUET, lineage, or_mychisq"
|
||||
, x = "Lineage"
|
||||
, y = "DUET (PS)")
|
||||
|
||||
#========================
|
||||
# Plot 4-5: Distributions
|
||||
# ggrdiges
|
||||
#========================
|
||||
|
||||
|
||||
#==================================================
|
||||
my_ats = 15 # axis text size
|
||||
my_als = 20 # axis label size
|
||||
|
||||
my_labels = c('Lineage 1', 'Lineage 2', 'Lineage 3', 'Lineage 4'
|
||||
#, 'Lineage 5', 'Lineage 6', 'Lineage 7'
|
||||
)
|
||||
names(my_labels) = c('lineage1', 'lineage2', 'lineage3', 'lineage4'
|
||||
# , 'lineage5', 'lineage6', 'lineage7'
|
||||
)
|
||||
|
||||
|
||||
#========================
|
||||
# Plot 4: Distribution
|
||||
# x = duet_scaled
|
||||
# y = country
|
||||
# fill = country_code
|
||||
# facet = lineage
|
||||
#========================
|
||||
# works neatly!
|
||||
|
||||
p1 = ggplot(df, aes(x = duet_scaled
|
||||
, y = country_code))+
|
||||
|
||||
#printFile=geom_density_ridges_gradient(
|
||||
geom_density_ridges_gradient(aes(fill = country_code)
|
||||
, jittered_points = TRUE
|
||||
, scale = 3
|
||||
, size = 0.3 ) +
|
||||
facet_wrap( ~lineage
|
||||
, scales = "free"
|
||||
, switch = 'x'
|
||||
, labeller = labeller(lineage = my_labels)
|
||||
) +
|
||||
coord_cartesian( xlim = c(-1, 1)) +
|
||||
#scale_fill_gradientn(colours = c("#f8766d", "white", "#00bfc4")
|
||||
# , name = "DUET" ) +
|
||||
theme(axis.text.x = element_text(size = my_ats
|
||||
, angle = 90
|
||||
, hjust = 1
|
||||
, vjust = 0.4)
|
||||
|
||||
#, axis.text.y = element_blank()
|
||||
, axis.title.x = element_blank()
|
||||
, axis.title.y = element_blank()
|
||||
, axis.ticks.y = element_blank()
|
||||
, plot.title = element_blank()
|
||||
, strip.text = element_text(size = my_als)
|
||||
, legend.text = element_text(size = my_als-5)
|
||||
, legend.title = element_text(size = my_als)
|
||||
)
|
||||
|
||||
p1
|
||||
|
||||
|
||||
#========================
|
||||
# Plot 5: Distribution
|
||||
# x = duet_scaled
|
||||
# y = country_code
|
||||
# fill = lineage
|
||||
# facet = NONE
|
||||
#========================
|
||||
# no facet wrap
|
||||
|
||||
p2 = ggplot(df, aes(x = duet_scaled
|
||||
, y = country_code))+
|
||||
|
||||
geom_density_ridges_gradient(aes(fill = factor(lineage))
|
||||
, scale = 3
|
||||
, size = 0.3 ) +
|
||||
coord_cartesian( xlim = c(-1, 1)) +
|
||||
#scale_fill_gradientn(colours = c("#f8766d", "white", "#00bfc4")
|
||||
# , name = "DUET" ) +
|
||||
#scale_fill_continuous(colours = c("darkgreen", "pink", "orange", "brown")
|
||||
# , name = "lineage" ) +
|
||||
theme(axis.text.x = element_text(size = my_ats
|
||||
, angle = 90
|
||||
, hjust = 1
|
||||
, vjust = 0.4)
|
||||
|
||||
#, axis.text.y = element_blank()
|
||||
, axis.title.x = element_blank()
|
||||
, axis.title.y = element_blank()
|
||||
, axis.ticks.y = element_blank()
|
||||
, plot.title = element_blank()
|
||||
, strip.text = element_text(size = my_als)
|
||||
, legend.text = element_text(size = my_als-5)
|
||||
, legend.title = element_text(size = my_als)
|
||||
)
|
||||
|
||||
p2
|
||||
|
||||
|
||||
#===============
|
||||
# lineage only
|
||||
#================
|
||||
#svg(plot_lineage_duet)
|
||||
p3 = ggplot(df, aes(x = duet_scaled
|
||||
, y = duet_outcome))+
|
||||
geom_density_ridges_gradient(aes(fill = ..x..)
|
||||
, jittered_points = TRUE
|
||||
, scale = 3
|
||||
, size = 0.3 ) +
|
||||
facet_wrap( ~lineage
|
||||
, scales = "free"
|
||||
#, switch = 'x'
|
||||
, labeller = labeller(lineage = my_labels) ) +
|
||||
coord_cartesian( xlim = c(-1, 1)) +
|
||||
scale_fill_gradientn(colours = c("#f8766d", "white", "#00bfc4")
|
||||
, name = "DUET" ) +
|
||||
theme(axis.text.x = element_text(size = my_ats
|
||||
, angle = 90
|
||||
, hjust = 1
|
||||
, vjust = 0.4)
|
||||
|
||||
, axis.text.y = element_blank()
|
||||
, axis.title.x = element_blank()
|
||||
, axis.title.y = element_blank()
|
||||
, axis.ticks.y = element_blank()
|
||||
, plot.title = element_blank()
|
||||
, strip.text = element_text(size = my_als)
|
||||
, legend.text = element_text(size = my_als-5)
|
||||
, legend.title = element_text(size = my_als)
|
||||
)
|
||||
|
||||
print(p3)
|
||||
Loading…
Add table
Add a link
Reference in a new issue