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

things

Browse source
This commit is contained in:
Tanushree Tunstall 2022-08-05 16:13:57 +01:00
parent 6cb9998c4c
commit 33925dafe9
4 changed files with 142 additions and 99 deletions
Download patch file Download diff file Expand all files Collapse all files

86
scripts/functions/dm_om_data.R
View file

@ -26,7 +26,8 @@ dm_om_wf_lf_data <- function(df
, categ_cols_to_factor){
df = as.data.frame(df)
df$maf = log10(df$maf) # can't see otherwise
# Initialise the required dfs based on gene name
geneL_normal = c("pnca")
geneL_na = c("gid", "rpob")
@ -93,7 +94,9 @@ dm_om_wf_lf_data <- function(df
, "mcsm_ppi2_affinity" , "mcsm_ppi2_scaled" , "mcsm_ppi2_outcome"
, "consurf_score" , "consurf_scaled" , "consurf_outcome" # exists now
, "snap2_score" , "snap2_scaled" , "snap2_outcome"
, "mcsm_na_affinity" , "mcsm_na_scaled" , "mcsm_na_outcome")
, "mcsm_na_affinity" , "mcsm_na_scaled" , "mcsm_na_outcome"
, "provean_score" , "provean_scaled" , "provean_outcome")
}else{
colnames_to_extract = c(mut_colname, mut_info_colname, mut_info_label_colname
, aa_pos_colname, LigDist_colname
@ -143,29 +146,31 @@ dynamut2_dn = paste0("Dynamut2 " , stability_suffix); dynamut2_dn
mcsm_na_dn = paste0("mCSM-NA ", stability_suffix); mcsm_na_dn
mcsm_ppi2_dn = paste0("mCSM-PPI2 ", stability_suffix); mcsm_ppi2_dn
consurf_dn = paste0("Consurf"); consurf_dn
consurf_dn = paste0("ConSurf"); consurf_dn
snap2_dn = paste0("SNAP2"); snap2_dn
provean_dn = paste0("PROVEAN"); provean_dn
# change column names: plyr
new_colnames = c(asa = "ASA"
, rsa = "RSA"
, rd_values = "RD"
, kd_values = "KD"
, log10_or_mychisq = "Log10 (OR)"
, neglog_pval_fisher = "-Log (P)"
#, af = "MAF"
, maf = "MAF"
#, ligand_dist_colname = lig_dn # cannot handle variable name 'ligand_dist_colname'
#, log10_or_mychisq = "Log10(OR)"
#, neglog_pval_fisher = "-Log(P)"
#, af = "MAF"
, maf = "Log10(MAF)"
#, ligand_dist_colname= lig_dn # cannot handle variable name 'ligand_dist_colname'
, affinity_scaled = mcsm_lig_dn
, duet_scaled = duet_dn
, foldx_scaled = foldx_dn
, deepddg_scaled = deepddg_dn
, ddg_dynamut2_scaled = dynamut2_dn
, mcsm_na_scaled = mcsm_na_dn
, mcsm_ppi2_affinity = mcsm_ppi2_dn
, consurf_score = consurf_dn
, snap2_score = snap2_dn)
, mcsm_ppi2_scaled = mcsm_ppi2_dn
, consurf_scaled = consurf_dn
, snap2_scaled = snap2_dn
, provean_scaled = provean_dn)
comb_df_sl1 = plyr::rename(comb_df_s
, replace = new_colnames
@ -173,7 +178,8 @@ comb_df_sl1 = plyr::rename(comb_df_s
, warn_duplicated = T)
# renaming colname using variable i.e ligand_dist_colname: dplyr
comb_df_sl = comb_df_sl1 %>% dplyr::rename(!!lig_dn := all_of(ligand_dist_colname))
#comb_df_sl = comb_df_sl1 %>% dplyr::rename(!!lig_dn := all_of(ligand_dist_colname))
comb_df_sl = comb_df_sl1 %>% dplyr::rename(!!lig_dn := all_of(LigDist_colname)) # NEW
names(comb_df_sl)
#=======================
@ -206,9 +212,9 @@ static_cols_end = c(lig_dn
, "RSA"
, "RD"
, "KD"
, "MAF"
, "Log10 (OR)"
#, "-Log (P)"
, "Log10(MAF)"
#, "Log10(OR)"
#, "-Log(P)"
)
#########################################################################
@ -344,19 +350,22 @@ wf_lf_dataL[['lf_dynamut2']] = lf_dynamut2
#1--> "most_variable", 2--> "", 3-->"", 4-->""
#5-->"", 6-->"", 7-->"", 8-->"", 9-->"most_conserved"
#====================
# FIXME: if you add category column to consurf
# WF data: consurf
cols_to_select_consurf = c(static_cols_start, c("consurf_outcome", consurf_dn), static_cols_end)
wf_consurf = comb_df_sl[, cols_to_select_consurf]
pivot_cols_consurf = cols_to_select_consurf[1: (length(static_cols_start) + 1)]; pivot_cols_consurf
# WF data: consurf
cols_to_select_consurf = c(static_cols_start, c(consurf_dn), static_cols_end)
wf_consurf = comb_df_sl[, cols_to_select_consurf]
pivot_cols_consurf = cols_to_select_consurf[1: (length(static_cols_start))]; pivot_cols_consurf
expected_rows_lf = nrow(wf_consurf) * (length(wf_consurf) - length(pivot_cols_consurf))
expected_rows_lf
# when outcome didn't exist
#cols_to_select_consurf = c(static_cols_start, c(consurf_dn), static_cols_end)
#wf_consurf = comb_df_sl[, cols_to_select_consurf]
#
# pivot_cols_consurf = cols_to_select_consurf[1: (length(static_cols_start))]; pivot_cols_consurf
# expected_rows_lf = nrow(wf_consurf) * (length(wf_consurf) - length(pivot_cols_consurf))
# expected_rows_lf
# LF data: consurf
lf_consurf = gather(wf_consurf
, key = param_type
@ -403,6 +412,37 @@ if (nrow(lf_snap2) == expected_rows_lf){
# Assign them to the output list
wf_lf_dataL[['wf_snap2']] = wf_snap2
wf_lf_dataL[['lf_snap2']] = lf_snap2
#==============
# Provean2: LF
#==============
# WF data: provean
cols_to_select_provean = c(static_cols_start, c("provean_outcome", provean_dn), static_cols_end)
wf_provean = comb_df_sl[, cols_to_select_provean]
pivot_cols_provean = cols_to_select_provean[1: (length(static_cols_start) + 1)]; pivot_cols_provean
expected_rows_lf = nrow(wf_provean) * (length(wf_provean) - length(pivot_cols_provean))
expected_rows_lf
# LF data: provean
lf_provean = gather(wf_provean
, key = param_type
, value = param_value
, all_of(provean_dn):tail(static_cols_end,1)
, factor_key = TRUE)
if (nrow(lf_provean) == expected_rows_lf){
cat("\nPASS: long format data created for ", provean_dn)
}else{
cat("\nFAIL: long format data could not be created for duet")
quit()
}
# Assign them to the output list
wf_lf_dataL[['wf_provean']] = wf_provean
wf_lf_dataL[['lf_provean']] = lf_provean
###########################################################################
# AFFINITY cols
###########################################################################