tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

agenerated corr plots with MAF and provean

Browse source
This commit is contained in:
Tanushree Tunstall 2022-08-06 13:08:42 +01:00
parent 1a513913ce
commit 569e372476
6 changed files with 115 additions and 104 deletions
Download patch file Download diff file Expand all files Collapse all files

26
scripts/Header_TT.R
View file

@ -223,20 +223,20 @@ consurf_palette2 = c("0" = "yellow2"
, "9" = "darkorchid4") , "9" = "darkorchid4")
consurf_colours = c(#"0" = rgb(1.00,1.00,0.59) consurf_colours = c(
"nsd" = rgb(1.00,1.00,0.59) "nsd" = rgb(1.00,1.00,0.59)
, "1" = rgb(0.63,0.16,0.37) , "1" = rgb(0.63,0.16,0.37)
, "2" = rgb(0.94,0.49,0.67) , "2" = rgb(0.94,0.49,0.67)
, "3" = rgb(0.98,0.78,0.86) , "3" = rgb(0.98,0.78,0.86)
, "4" = rgb(0.98,0.92,0.96) , "4" = rgb(0.98,0.92,0.96)
, "5" = rgb(1.00,1.00,1.00) , "5" = rgb(1.00,1.00,1.00)
, "6" = rgb(0.84,0.94,0.94) , "6" = rgb(0.84,0.94,0.94)
, "7" = rgb(0.65,0.86,0.90) , "7" = rgb(0.65,0.86,0.90)
, "8" = rgb(0.29,0.69,0.75) , "8" = rgb(0.29,0.69,0.75)
, "9" = rgb(0.04,0.49,0.51) , "9" = rgb(0.04,0.49,0.51)
) )
consurf_bp_colours = c(#"0" = rgb(1.00,1.00,0.59) consurf_bp_colours = c(
"0" = rgb(1.00,1.00,0.59) "0" = rgb(1.00,1.00,0.59)
, "1" = rgb(0.63,0.16,0.37) , "1" = rgb(0.63,0.16,0.37)
, "2" = rgb(0.94,0.49,0.67) , "2" = rgb(0.94,0.49,0.67)

120
scripts/functions/corr_plot_data.R
View file

@ -22,9 +22,12 @@
#1) Corr type? #1) Corr type?
#2) #2)
################################################################## ##################################################################
# LigDist_colname #from globals: plotting_globals.R
# ppi2Dist_colname #from globals: plotting_globals.R
# naDist_colname #from globals: plotting_globals.R
corr_data_extract <- function(df corr_data_extract <- function(df
#, gene_name = gene , gene
, drug_name = drug , drug
#, ligand_dist_colname = LigDist_colname #, ligand_dist_colname = LigDist_colname
, colnames_to_extract , colnames_to_extract
, colnames_display_key , colnames_display_key
@ -38,34 +41,48 @@ corr_data_extract <- function(df
, "\n=========================================") , "\n=========================================")
cat("\nExtracting default columns for" cat("\nExtracting default columns for"
#, "\nGene name:", gene , "\nGene name:", gene
, "\nDrug name:", drug) , "\nDrug name:", drug)
colnames_to_extract = c(drug geneL_normal = c("pnca")
#, "mutationinformation" geneL_na = c("gid", "rpob")
#, "mutation_info_labels" geneL_ppi2 = c("alr", "embb", "katg", "rpob")
, "dst_mode"
, "duet_stability_change" common_colnames = c(drug, "dst_mode"
, "ligand_affinity_change" , "duet_stability_change" , "ddg_foldx" , "deepddg" , "ddg_dynamut2"
, "ligand_distance" , "asa" , "rsa" , "kd_values" , "rd_values"
#, ligand_dist_colname , "maf" , "log10_or_mychisq" , "neglog_pval_fisher"
, "interface_dist" , LigDist_colname
, "ddg_foldx" , "consurf_score" , "snap2_score" , "provean_score"
, "deepddg" , "ligand_affinity_change"
, "asa" #, "ddg_dynamut", "ddg_encom", "dds_encom", "ddg_mcsm", "ddg_sdm", "ddg_duet"
, "rsa" )
, "kd_values"
, "rd_values" display_common_colnames = c( drug, "dst_mode"
, "af" , "DUET" , "FoldX" , "DeepDDG", "Dynamut2"
, "log10_or_mychisq" , "ASA" , "RSA" , "KD" , "RD"
, "neglog_pval_fisher" , "MAF" , "Log(OR)" , "-Log(P)"
, "ddg_dynamut2" , "Lig-Dist"
, "consurf_score" , "ConSurf" , "SNAP2" , "PROVEAN"
, "snap2_score" , "mCSM-lig"
, "ddg_dynamut", "ddg_encom", "dds_encom", "ddg_mcsm", "ddg_sdm", "ddg_duet" # , "Dynamut" , "ENCoM-DDG" , "mCSM" , "SDM" , "DUET-d" , "ENCoM-DDS"
, "mcsm_na_affinity" )
, "mcsm_ppi2_affinity"
) if (tolower(gene)%in%geneL_normal){
colnames_to_extract = c(common_colnames)
display_colnames = c(display_common_colnames)
}
if (tolower(gene)%in%geneL_ppi2){
colnames_to_extract = c(common_colnames ,"mcsm_ppi2_affinity", ppi2Dist_colname)
display_colnames = c(display_common_colnames,"mCSM-PPI2" , "PPI-Dist")
}
if (tolower(gene)%in%geneL_na){
colnames_to_extract = c(common_colnames,"mcsm_na_affinity", naDist_colname)
display_colnames = c(display_common_colnames, "mCSM-NA", "NA-Dist")
}
# [optional] arg: extract_scaled_cols # [optional] arg: extract_scaled_cols
if (extract_scaled_cols){ if (extract_scaled_cols){
@ -77,46 +94,15 @@ corr_data_extract <- function(df
colnames_to_extract = colnames_to_extract colnames_to_extract = colnames_to_extract
} }
corr_df = df[, colnames(df)%in%colnames_to_extract] # extract df based on gene
corr_df = df[,colnames_to_extract]
colnames(corr_df)
display_colnames
# arg: colnames_display_key # arg: colnames_display_key
colnames_display_key = c(duet_stability_change = "DUET" colnames(corr_df)[colnames(corr_df)%in%colnames_to_extract] <- display_colnames
, ligand_affinity_change = "mCSM-lig" colnames(corr_df)
, ligand_distance = "ligand_distance"
#, ligand_dist_colname = "ligand_distance"
, interface_dist = "interface_dist"
, ddg_foldx = "FoldX"
, deepddg = "DeepDDG"
, asa = "ASA"
, rsa = "RSA"
, kd_values = "KD"
, rd_values = "RD"
, af = "MAF"
, log10_or_mychisq = "Log (OR)"
, neglog_pval_fisher = "-Log (P)"
, ddg_dynamut2 = "Dynamut2"
, consurf_score = "Consurf"
, snap2_score = "SNAP2"
, ddg_dynamut = "Dynamut"
, ddg_encom = "ENCoM-DDG"
, ddg_mcsm = "mCSM"
, ddg_sdm = "SDM"
, ddg_duet = "DUET-d"
, dds_encom = "ENCoM-DDS"
, mcsm_na_affinity = "mCSM-NA"
, mcsm_ppi2_affinity = "mCSM-PPI2")
# COMMENT: This only works when all the columns are in the namekey vector.
# If one is missing, there is no error, but it also renamed as "NA.
#names(corr_df) <- colnames_display_key[names(corr_df)]
# Solution: to use plyr::rename()
# Consider using requireNamespace() instead of library() so its function names doesn't collide with dplyr's.
corr_df = plyr::rename(corr_df
, replace = colnames_display_key
, warn_missing = T
, warn_duplicated = T)
cat("\nExtracted ncols:", ncol(corr_df) cat("\nExtracted ncols:", ncol(corr_df)
,"\nRenaming successful") ,"\nRenaming successful")

6
scripts/functions/plotting_globals.R
View file

@ -39,9 +39,9 @@ resistance_col <<- "drtype"
LigDist_colname <<- "ligand_distance" LigDist_colname <<- "ligand_distance"
LigDist_cutoff <<- 10 LigDist_cutoff <<- 10
DistCutOff = 10 DistCutOff <<- 10
ppi2Dist_colname = "interface_dist" ppi2Dist_colname <<- "interface_dist"
naDist_colname = "TBC" naDist_colname <<- "TBC"
#================== #==================
# Angstroms symbol # Angstroms symbol

10
scripts/plotting/get_plotting_dfs.R
View file

@ -186,10 +186,16 @@ cat(s3)
# make sure the above script works because merged_df2_combined is needed # make sure the above script works because merged_df2_combined is needed
merged_df3 = as.data.frame(merged_df3) merged_df3 = as.data.frame(merged_df3)
corr_df_m3_f = corr_data_extract(merged_df3, extract_scaled_cols = F) corr_df_m3_f = corr_data_extract(merged_df3
, gene = gene
, drug = drug
, extract_scaled_cols = F)
head(corr_df_m3_f) head(corr_df_m3_f)
corr_df_m2_f = corr_data_extract(merged_df2, extract_scaled_cols = F) corr_df_m2_f = corr_data_extract(merged_df2
, gene = gene
, drug = drug
, extract_scaled_cols = F)
head(corr_df_m2_f) head(corr_df_m2_f)
s4 = c("\nSuccessfully sourced Corr_data.R") s4 = c("\nSuccessfully sourced Corr_data.R")

51
scripts/plotting/plotting_thesis/corr_plots_thesis.R
View file

@ -1,5 +1,9 @@
merged_df3 = as.data.frame(merged_df3) merged_df3 = as.data.frame(merged_df3)
corr_plotdf = corr_data_extract(merged_df3, extract_scaled_cols = F) corr_plotdf = corr_data_extract(merged_df3
, gene = gene
, drug = drug
, extract_scaled_cols = F)
colnames(corr_plotdf)
#================ #================
# stability # stability
@ -9,12 +13,13 @@ corr_ps_colnames = c("DUET"
, "DeepDDG" , "DeepDDG"
, "Dynamut2" , "Dynamut2"
, "MAF" , "MAF"
, "Log (OR)" , "Log(OR)"
, "-Log (P)" , "-Log(P)"
#, "ligand_distance" #, "ligand_distance"
, "dst_mode" , "dst_mode"
, drug) , drug)
corr_ps_colnames%in%colnames(corr_plotdf)
corr_df_ps = corr_plotdf[, corr_ps_colnames] corr_df_ps = corr_plotdf[, corr_ps_colnames]
color_coln = which(colnames(corr_df_ps) == "dst_mode") color_coln = which(colnames(corr_df_ps) == "dst_mode")
@ -46,10 +51,10 @@ my_corr_pairs(corr_data_all = corr_df_ps
dev.off() dev.off()
##################################################### #####################################################
DistCutOff = 10 #DistCutOff = 10
LigDist_colname # = "ligand_distance" # from globals #LigDist_colname # = "ligand_distance" # from globals
ppi2Dist_colname = "interface_dist" #ppi2Dist_colname = "interface_dist"
naDist_colname = "TBC" #naDist_colname = "TBC"
##################################################### #####################################################
#================ #================
@ -57,14 +62,15 @@ naDist_colname = "TBC"
#================ #================
corr_lig_colnames = c("mCSM-lig" corr_lig_colnames = c("mCSM-lig"
, "MAF" , "MAF"
, "Log (OR)" , "Log(OR)"
, "-Log (P)" , "-Log(P)"
, "ligand_distance" , "Lig-Dist"
, "dst_mode" , "dst_mode"
, drug) , drug)
corr_lig_colnames%in%colnames(corr_plotdf)
corr_df_lig = corr_plotdf[, corr_lig_colnames] corr_df_lig = corr_plotdf[, corr_lig_colnames]
corr_df_lig = corr_df_lig[corr_df_lig[[LigDist_colname]]<DistCutOff,] corr_df_lig = corr_df_lig[corr_df_lig["Lig-Dist"]<DistCutOff,]
color_coln = which(colnames(corr_df_lig) == "dst_mode") color_coln = which(colnames(corr_df_lig) == "dst_mode")
end = which(colnames(corr_df_lig) == drug) end = which(colnames(corr_df_lig) == drug)
@ -99,15 +105,15 @@ dev.off()
#================ #================
corr_ppi2_colnames = c("mCSM-PPI2" corr_ppi2_colnames = c("mCSM-PPI2"
, "MAF" , "MAF"
, "Log (OR)" , "Log(OR)"
, "-Log (P)" , "-Log(P)"
, "interface_dist" , "PPI-Dist" # "interface_dist"
, "dst_mode" , "dst_mode"
, drug) , drug)
corr_ppi2_colnames%in%colnames(corr_plotdf)
corr_df_ppi2 = corr_plotdf[, corr_ppi2_colnames] corr_df_ppi2 = corr_plotdf[, corr_ppi2_colnames]
corr_df_ppi2 = corr_df_ppi2[corr_df_ppi2[[ppi2Dist_colname]]<DistCutOff,] corr_df_ppi2 = corr_df_ppi2[corr_df_ppi2["PPI-Dist"]<DistCutOff,]
color_coln = which(colnames(corr_df_ppi2) == "dst_mode") color_coln = which(colnames(corr_df_ppi2) == "dst_mode")
end = which(colnames(corr_df_ppi2) == drug) end = which(colnames(corr_df_ppi2) == drug)
@ -146,10 +152,21 @@ dev.off()
# FIXME: ADD PROVEAN
#################################################### ####################################################
# CONSERVATION # CONSERVATION
corr_conservation_cols = c("ConSurf"
, "SNAP2"
, "PROVEAN"
, "MAF"
, "Log(OR)"
, "-Log(P)"
, "dst_mode"
, drug)
#################################################### ####################################################
colnames(corr_plotdf)
corr_conservation_cols%in%colnames(corr_plotdf)
corr_df_cons = corr_plotdf[, corr_conservation_cols] corr_df_cons = corr_plotdf[, corr_conservation_cols]
color_coln = which(colnames(corr_df_cons) == "dst_mode") color_coln = which(colnames(corr_df_cons) == "dst_mode")

6
scripts/plotting/plotting_thesis/preformatting.R
View file

@ -111,7 +111,9 @@ df3 = merged_df3
#df3[[consurf_colNew]] = as.factor(df3[[consurf_colNew]]) #df3[[consurf_colNew]] = as.factor(df3[[consurf_colNew]])
#df3[[consurf_colNew]] #df3[[consurf_colNew]]
# not this bit # not this bit
levels(df3$consurf_outcome) = c( "nsd", 1, 2, 3, 4, 5, 6, 7, 8, 9) #!!!!!!!!!!!!!1
#levels(df3$consurf_outcome) = c( "nsd", 1, 2, 3, 4, 5, 6, 7, 8, 9)
#levels(df3$consurf_outcome) #levels(df3$consurf_outcome)
# SNAP2 labels # SNAP2 labels
@ -242,9 +244,9 @@ corr_ppi2_colnames = c("mCSM-PPI2"
, "dst_mode" , "dst_mode"
, drug) , drug)
#FIXME: Add provean
corr_conservation_cols = c("Consurf" corr_conservation_cols = c("Consurf"
, "SNAP2" , "SNAP2"
, "PROVEAN"
, "MAF" , "MAF"
, "Log (OR)" , "Log (OR)"
, "-Log (P)" , "-Log (P)"