more plot modifications dm and om plots mainly

This commit is contained in:
Tanushree Tunstall 2022-08-08 15:32:16 +01:00
parent 4e6f10d1ba
commit 0234a8f77b
6 changed files with 501 additions and 358 deletions

View file

@ -1,237 +1,230 @@
#################
# Numbers
##################
nrow(wf_mcsm_lig)
table(wf_mcsm_lig$mutation_info_labels)
nrow(wf_mcsm_ppi2)
table(wf_mcsm_ppi2$mutation_info_labels)
all_dm_om_df = dm_om_wf_lf_data(df = merged_df3, gene = gene)
#
# lf_duet = all_dm_om_df[['lf_duet']]
# table(lf_duet$param_type)
################################################################
geneL_normal = c("pnca")
geneL_na = c("gid", "rpob")
geneL_ppi2 = c("alr", "embb", "katg", "rpob")
#======================
# Data: Dist+genomics
#======================
lf_dist_genP = all_dm_om_df[['lf_dist_gen']]
wf_dist_genP = all_dm_om_df[['wf_dist_gen']]
levels(lf_dist_genP$param_type)
#lf_dist_genP$param_type <- factor(lf_dist_genP$param_type, levels=c("Log10(MAF)", "Lig Dist(Å)", "PPI Dist(Å)"))
table(lf_dist_genP$param_type)
if (tolower(gene)%in%geneL_na){
lf_mcsm_na
}
genomics_param = c("Log10(MAF)")
if (tolower(gene)%in%geneL_ppi2){
lf_mcsm_ppi2
}
dist_genP = lf_bp2(lf_dist_genP
#, p_title
, violin_quantiles = c(0.5), monochrome = F)
colnames(lf_duet)
table(lf_duet$param_type)
#-------------------
# Genomics data plot
#-------------------
genomics_dataP = lf_dist_genP[lf_dist_genP$param_type%in%genomics_param,]
genomics_dataP$param_type = factor(genomics_dataP$param_type)
table(genomics_dataP$param_type)
static_colsP = c("Lig Dist(Å)","ASA", "RSA","RD","KD","Log10(MAF)")
genomicsP = lf_bp2(genomics_dataP
#, p_title = ""
, violin_quantiles = c(0.5), monochrome = F)
stability_suffix <- paste0(delta_symbol, delta_symbol, "G")
lf_commonP = lf_duet[!lf_duet$param_type%in%c("DUET ΔΔG"),]
lf_commonP$param_type = levels(droplevels(lf_commonP$param_type))
table(lf_commonP$param_type); colnames(lf_commonP)
lf_commonP$outcome = lf_commonP$duet_outcome
lf_commonP$duet_outcome = NULL
genomicsP
#check
wilcox.test(wf_dist_genP$`Log10(MAF)`[wf_dist_genP$mutation_info_labels=="R"]
, wf_dist_genP$`Log10(MAF)`[wf_dist_genP$mutation_info_labels=="S"], paired = FALSE)
lf_duet$outcome = lf_duet$duet_outcome
lf_duet$duet_outcome = NULL
lf_duetP = lf_duet[!lf_duet$param_type%in%c(static_colsP, "outcome"),]
lf_duetP$param_type = levels(droplevels(lf_duetP$param_type))
table(lf_duetP$param_type); colnames(lf_duetP)
colnames(lf_duetP)
tapply(wf_dist_genP$`Log10(MAF)`, wf_dist_genP$mutation_info_labels, summary)
lf_foldx$outcome = lf_foldx$foldx_outcome
lf_foldx$foldx_outcome = NULL
lf_foldxP = lf_foldx[!lf_foldx$param_type%in%c(static_colsP,"outcome"),]
lf_foldxP$param_type = levels(droplevels(lf_foldxP$param_type))
table(lf_foldxP$param_type)
colnames(lf_foldxP)
#-------------------
# Distance data plot: not genomics data
#-------------------
dist_dataP = lf_dist_genP[!lf_dist_genP$param_type%in%genomics_param,]
#dist_dataP$param_type = factor(dist_dataP$param_type)
table(dist_dataP$param_type)
distanceP = lf_bp2(dist_dataP
#, p_title = ""
, violin_quantiles = c(0.5), monochrome = F)
lf_deepddg$outcome = lf_deepddg$deepddg_outcome
lf_deepddg$deepddg_outcome = NULL
lf_deepddgP = lf_deepddg[!lf_deepddg$param_type%in%c(static_colsP, "outcome"),]
lf_deepddgP$param_type = levels(droplevels(lf_deepddgP$param_type))
distanceP
# check
wilcox.test(wf_dist_genP$`PPI Dist(Å)`[wf_dist_genP$mutation_info_labels=="R"]
, wf_dist_genP$`PPI Dist(Å)`[wf_dist_genP$mutation_info_labels=="S"], paired = FALSE)
wilcox.test(wf_dist_genP$`Lig Dist(Å)`[wf_dist_genP$mutation_info_labels=="R"]
, wf_dist_genP$`Lig Dist(Å)`[wf_dist_genP$mutation_info_labels=="S"], paired = FALSE)
tapply(wf_dist_genP$`PPI Dist(Å)`, wf_dist_genP$mutation_info_labels, summary)
tapply(wf_dist_genP$`Lig Dist(Å)`, wf_dist_genP$mutation_info_labels, summary)
#==============
# Plot:DUET
#==============
lf_duetP = all_dm_om_df[['lf_duet']]
#lf_duetP = lf_duet[!lf_duet$param_type%in%c(static_colsP),]
table(lf_duetP$param_type)
lf_duetP$param_type = factor(lf_duetP$param_type)
table(lf_duetP$param_type)
duetP = lf_bp2(lf_duetP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
#==============
# Plot:FoldX
#==============
lf_foldxP = all_dm_om_df[['lf_foldx']]
#lf_foldxP = lf_foldx[!lf_foldx$param_type%in%c(static_colsP),]
table(lf_foldxP$param_type)
lf_foldxP$param_type = factor(lf_foldxP$param_type)
table(lf_foldxP$param_type)
foldxP = lf_bp2(lf_foldxP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
#==============
# Plot:DeepDDG
#==============
lf_deepddgP = all_dm_om_df[['lf_deepddg']]
#lf_deepddgP = lf_deepddg[!lf_deepddg$param_type%in%c(static_colsP),]
table(lf_deepddgP$param_type)
lf_deepddgP$param_type = factor(lf_deepddgP$param_type)
table(lf_deepddgP$param_type)
colnames(lf_deepddgP)
deepddgP = lf_bp2(lf_deepddgP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F
, dot_transparency = 0.3)
deepddgP
lf_dynamut2$outcome = lf_dynamut2$ddg_dynamut2_outcome
lf_dynamut2$ddg_dynamut2_outcome = NULL
lf_dynamut2P = lf_dynamut2[!lf_dynamut2$param_type%in%c(static_colsP, "outcome"),]
lf_dynamut2P$param_type = levels(droplevels(lf_dynamut2P$param_type))
#==============
# Plot: Dynamut2
#==============
lf_dynamut2P = all_dm_om_df[['lf_dynamut2']]
#lf_dynamut2P = lf_dynamut2[!lf_dynamut2$param_type%in%c(static_colsP),]
table(lf_dynamut2P$param_type)
lf_dynamut2P$param_type = factor(lf_dynamut2P$param_type)
table(lf_dynamut2P$param_type)
colnames(lf_dynamut2P)
dynamut2P = lf_bp2(lf_dynamut2P
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
lf_consurf$outcome = lf_consurf$consurf_outcome
lf_consurf$consurf_outcome = NULL
lf_consurfP = lf_consurf[!lf_consurf$param_type%in%c(static_colsP),]
lf_consurfP$param_type = levels(droplevels(lf_consurfP$param_type))
#==============
# Plot:ConSurf
#==============
lf_consurfP = all_dm_om_df[['lf_consurf']]
#lf_consurfP = lf_consurf[!lf_consurf$param_type%in%c(static_colsP),]
table(lf_consurfP$param_type)
lf_consurfP$param_type = factor(lf_consurfP$param_type)
table(lf_consurfP$param_type)
colnames(lf_consurfP)
consurfP = lf_bp2(lf_consurfP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
lf_snap2$outcome = lf_snap2$snap2_outcome
lf_snap2$snap2_outcome = NULL
lf_snap2P = lf_snap2[!lf_snap2$param_type%in%c(static_colsP),]
lf_snap2P$param_type = levels(droplevels(lf_snap2P$param_type))
#==============
# Plot:SNAP2
#==============
lf_snap2P = all_dm_om_df[['lf_snap2']]
#lf_snap2P = lf_snap2[!lf_snap2$param_type%in%c(static_colsP),]
table(lf_snap2P$param_type)
lf_snap2P$param_type = factor(lf_snap2P$param_type)
table(lf_snap2P$param_type)
colnames(lf_snap2P)
snap2P = lf_bp2(lf_snap2P
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
lf_provean$outcome = lf_provean$provean_outcome
lf_provean$provean_outcome = NULL
lf_proveanP = lf_provean[!lf_provean$param_type%in%c(static_colsP),]
lf_proveanP$param_type = levels(droplevels(lf_proveanP$param_type))
#==============
# Plot:PROVEAN
#==============
lf_proveanP = all_dm_om_df[['lf_provean']]
#lf_proveanP = lf_provean[!lf_provean$param_type%in%c(static_colsP),]
table(lf_proveanP$param_type)
lf_proveanP$param_type = factor(lf_proveanP$param_type)
table(lf_proveanP$param_type)
colnames(lf_proveanP)
bar = rbind(colnames(lf_duetP)
, colnames(lf_foldxP)
, colnames(lf_deepddgP)
, colnames(lf_dynamut2P)
, colnames(lf_consurfP)
, colnames(lf_snap2P)
, colnames(lf_proveanP)
)
bar
lf_df_stabP = rbind((lf_duetP)
, (lf_foldxP)
, (lf_deepddgP)
, (lf_dynamut2P))
lf_df_consP = rbind((lf_consurfP)
, (lf_snap2P)
, (lf_proveanP))
table(lf_df_stabP$param_type)
# VERY USEFUL for seeing numbers for param types
table(lf_df_stabP$param_type,lf_df_stabP$outcome)
table(lf_df_consP$param_type,lf_df_consP$outcome)
#==============
# Plot:BP
#==============
stability_suffix <- paste0(delta_symbol, delta_symbol, "G")
# lf_bp(lf_df_stabP, p_title = paste0("Stability",stability_suffix)
# , violin_quantiles = c(0.5))
# lf_bp(lf_df_consP, p_title = "Evolutionary Conservation"
# , violin_quantiles = c(0.5))
lf_bp2(lf_df_stabP, p_title = paste0("Stability",stability_suffix)
proveanP = lf_bp2(lf_proveanP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
lf_bp2(lf_duet, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
lf_bp2(lf_df_consP, p_title = "Evolutionary Conservation"
, violin_quantiles = c(0.5), monochrome = F)
#HMMM: Bollocks!
lf_bp2(lf_commonP, p_title = paste0("Residue level properties")
, violin_quantiles = c(0.5)
, monochrome = T) # doesn't plot stat bars
lf_bp(lf_commonP, p_title = paste0("Residue level properties")
, violin_quantiles = c(0.5)) #plots stat bars but incorrect result
lf_unpaired_stats(lf_duet)
wilcox.test(wf_duet$`Lig Dist(Å)`[wf_duet$mutation_info_labels=="R"]
, wf_duet$`Lig Dist(Å)`[wf_duet$mutation_info_labels=="S"])
wilcox.test(wf_duet$ASA[wf_duet$mutation_info_labels=="R"]
, wf_duet$ASA[wf_duet$mutation_info_labels=="S"])
# 1: variable
# 9: conserved
# CHECK THESE
foo = merged_df3[c("dst_mode", "mutation_info_labels", "consurf_colour_rev"
, "consurf_scaled"
, "consurf_score"
, "consurf_outcome"
, "snap2_score"
, "snap2_scaled"
, "snap2_outcome"
, "provean_score"
, "provean_scaled"
, "provean_outcome")]
################
# Affinity
################
# ligand
lf_mcsm_lig$outcome = lf_mcsm_lig$ligand_outcome
lf_mcsm_lig$ligand_outcome = NULL
colnames(lf_mcsm_lig)
table(lf_mcsm_lig$param_type)
lf_mcsm_lig$outcome = lf_mcsm_lig$ligand_outcome
lf_mcsm_ligP = lf_mcsm_lig[!lf_mcsm_lig$param_type%in%c(static_colsP, "outcome"),]
#lf_mcsm_ligP$param_type = levels(droplevels(lf_mcsm_ligP))
#==============
# Plot: mCSM-lig
#==============
lf_mcsm_ligP = all_dm_om_df[['lf_mcsm_lig']]
#lf_mcsm_ligP = lf_mcsm_lig[!lf_mcsm_lig$param_type%in%c(static_colsP),]
table(lf_mcsm_ligP$param_type)
lf_mcsm_ligP$param_type = factor(lf_mcsm_ligP$param_type)
table(lf_mcsm_ligP$param_type)
lf_mcsm_ligP$ligand_outcome = NULL
colnames(lf_mcsm_ligP)
if (tolower(gene)%in%geneL_na){
lf_mcsm_na$outcome = lf_mcsm_na$mcsm_na_outcome
#lf_mcsm_na$mcsm_na_outcome = NULL
lf_mcsm_naP = lf_mcsm_na[!lf_mcsm_na$param_type%in%c(static_colsP, "outcome"),]
#lf_mcsm_naP$param_type = levels(droplevels(lf_mcsm_naP))
table(lf_mcsm_naP$param_type)
lf_mcsm_naP$mcsm_na_outcome = NULL
colnames(lf_mcsm_naP)
}
mcsmligP = lf_bp2(lf_mcsm_ligP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
#==============
# Plot: mCSM-ppi2
#==============
if (tolower(gene)%in%geneL_ppi2){
lf_mcsm_ppi2$outcome = lf_mcsm_ppi2$mcsm_ppi2_outcome
colnames(lf_mcsm_ppi2)
#lf_mcsm_ppi2$mcsm_ppi2_outcome = NULL
lf_mcsm_ppi2P = lf_mcsm_ppi2[!lf_mcsm_ppi2$param_type%in%c(static_colsP, "outcome"),]
#lf_mcsm_ppi2P$param_type = levels(droplevels(lf_mcsm_ppi2P))
lf_mcsm_ppi2P = all_dm_om_df[['lf_mcsm_ppi2']]
#lf_mcsm_ppi2P = lf_mcsm_ppi2[!lf_mcsm_ppi2$param_type%in%c(static_colsP),]
table(lf_mcsm_ppi2P$param_type)
lf_mcsm_ppi2P$param_type = factor(lf_mcsm_ppi2P$param_type)
table(lf_mcsm_ppi2P$param_type)
lf_mcsm_ppi2P$mcsm_ppi2_outcome = NULL
colnames(lf_mcsm_ppi2P)
mcsmppi2P = lf_bp2(lf_mcsm_ppi2P
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
}
#==============
# Plot: mCSM-NA
#==============
if (tolower(gene)%in%geneL_na){
lf_mcsm_naP = all_dm_om_df[['lf_mcsm_na']]
#lf_mcsm_naP = lf_mcsm_na[!lf_mcsm_na$param_type%in%c(static_colsP),]
table(lf_mcsm_naP$param_type)
lf_mcsm_naP$param_type = factor(lf_mcsm_naP$param_type)
table(lf_mcsm_naP$param_type)
mcsmnaP = lf_bp2(lf_mcsm_naP
#, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
}
bar = rbind(colnames(lf_mcsm_ligP)
#, colnames(lf_mcsm_naP)
, colnames(lf_mcsm_ppi2P))
bar
######################################
# Outplot with stats
######################################
lf_df_affP = rbind((lf_mcsm_ligP)
, (lf_mcsm_ppi2P))
cowplot::plot_grid(
cowplot::plot_grid(duetP, foldxP, deepddgP, dynamut2P, genomicsP, distanceP
, nrow=1),
# cowplot::plot_grid(genomicsP, distanceP
# , nrow = 1),
cowplot::plot_grid(consurfP, snap2P, proveanP
, mcsmligP
, mcsmppi2P
#, mcsmnaP
, nrow=1),
nrow=2)
lf_bp(lf_df_affP, p_title = paste0("Affinity changes")
, violin_quantiles = c(0.5))
#, monochrome = T) # doesn't plot stat bars
wilcox.test(wf_mcsm_lig$ASA[wf_mcsm_lig$mutation_info_labels=="R"]
, wf_mcsm_lig$ASA[wf_mcsm_lig$mutation_info_labels=="S"])
foo = lf_consurfP
wilcox.test(wf_mcsm_ppi2$ASA[wf_mcsm_ppi2$mutation_info_labels=="R"]
, wf_mcsm_ppi2$ASA[wf_mcsm_ppi2$mutation_info_labels=="S"])
# proveanP = lf_bp2(lf_proveanP, colour_categ = "mutation_info_labels"
# , p_title = paste0("Evolutionary conservation")
# , dot_transparency = 1
# , violin_quantiles = c(0.5), monochrome = F)
#
# proveanP
#===============================
p1 = lf_bp2(lf_df_stabP, p_title = paste0("Stability",stability_suffix)
, violin_quantiles = c(0.5), monochrome = F)
p2 = lf_bp2(lf_df_consP, p_title = "Evolutionary Conservation"
, violin_quantiles = c(0.5), monochrome = F)
p3 = lf_bp2(lf_df_affP, p_title = paste0("Affinity changes")
, violin_quantiles = c(0.5), monochrome = F)
cowplot::plot_grid(p1,cowplot::plot_grid(p2, p3), nrow=2)