In sbenfatto/PARIS: PAn-canceR Inferred Synthetic lethalities (PARIS), a machine learning-based approach to predict cancer vulnerabilities

knitr::opts_chunk$set(
  collapse = TRUE,
  comment = "#>"
)

library(PARIS)

library(ggrepel) library(openxlsx)

select_pairs("/mnt/data/DepMap19Q3/Models_DDR_vs_DDR/x_paper/", coeff.filter = T) -> mydata_DDR select_pairs("/mnt/data/DepMap19Q3/Models_DDR_vs_ALL/x_paper/", coeff.filter = T) -> mydata_ALL

write files

file_selected_pairs_DDR <- "/mnt/data/DepMap19Q3/Models_DDR_vs_DDR/x_paper/Selected_genes_DDR_vs_DDR.xlsx" wb <- createWorkbook() addWorksheet(wb, "er") addWorksheet(wb, "eg") addWorksheet(wb, "ec") addWorksheet(wb, "mr") addWorksheet(wb, "mg") addWorksheet(wb, "mc")

writeData(wb, "er", filter(mydata_DDR, group == "er"), startRow = 1, startCol = 1) writeData(wb, "eg", filter(mydata_DDR, group == "eg"), startRow = 1, startCol = 1) writeData(wb, "ec", filter(mydata_DDR, group == "ec"), startRow = 1, startCol = 1) writeData(wb, "mr", filter(mydata_DDR, group == "mr"), startRow = 1, startCol = 1) writeData(wb, "mg", filter(mydata_DDR, group == "mg"), startRow = 1, startCol = 1) writeData(wb, "mc", filter(mydata_DDR, group == "mc"), startRow = 1, startCol = 1)

saveWorkbook(wb, file = file_selected_pairs_DDR)

file_selected_pairs_ALL <- "/mnt/data/DepMap19Q3/Models_DDR_vs_ALL/x_paper/Selected_genes_DDR_vs_ALL.xlsx" wb <- createWorkbook() addWorksheet(wb, "er") addWorksheet(wb, "eg") addWorksheet(wb, "ec") addWorksheet(wb, "mr") addWorksheet(wb, "mg") addWorksheet(wb, "mc")

writeData(wb, "er", filter(mydata_ALL, group == "er"), startRow = 1, startCol = 1) writeData(wb, "eg", filter(mydata_ALL, group == "eg"), startRow = 1, startCol = 1) writeData(wb, "ec", filter(mydata_ALL, group == "ec"), startRow = 1, startCol = 1) writeData(wb, "mr", filter(mydata_ALL, group == "mr"), startRow = 1, startCol = 1) writeData(wb, "mg", filter(mydata_ALL, group == "mg"), startRow = 1, startCol = 1) writeData(wb, "mc", filter(mydata_ALL, group == "mc"), startRow = 1, startCol = 1)

saveWorkbook(wb, file = file_selected_pairs_ALL)

ANALYSIS STRINGDB

subset by threshold

mydata_DDR_sub <- subset_by_threshold(mydata_DDR, "e", 0.4)

retrieve stringdb interaction

mydata_DDR_sub_string <- get_stringdb_interactions(mydata_DDR_sub) mydata_DDR_sub_string_freq <- compute_interaction_freq(mydata_DDR_sub, mydata_DDR_sub_string)

plot combined score

plot.stringdb.interactions.combined.score.by.group(mydata_DDR_sub_string)

plot freq of interaction in selected pairs by group

plot.stringdb.interactions.by.group(mydata_DDR_sub_string_freq)

compare DDR vs DDR and DDR vs ALL commonly by group

inner_join(mydata_DDR, mydata_ALL, by = c("gene", "c", "group")) %>% ggplot(aes(scaled.x, scaled.y))+geom_point()+ facet_wrap(~group)+geom_smooth(method = "lm")+ stat_cor()+ xlab("scaled importance DDR vs DDR")+ylab("scaled importance DDR vs ALL")+theme_bw()+ theme(text = element_text(size = 20), axis.text = element_text(colour = "black"))+ geom_label_repel(data = . %>% mutate(labels=ifelse((scaled.x > 0.4 | scaled.y > 0.4), as.character(interaction(gene,c)), "")), aes(label=labels), box.padding = 1)

PARALOGS ANALYSIS

paralogsDDR <- get_paralog_biomart(unique(c(mydata_DDR$gene))) %>% as_tibble() %>% filter(hsapiens_paralog_associated_gene_name != "") colnames(paralogsDDR) <- c("gene", "c") paralogsDDR$paralogs <- TRUE

left_join(mydata_DDR, paralogsDDR) %>% replace_na(list(paralogs = "FALSE")) %>% ggplot(aes(paralogs, scaled))+ geom_boxplot(outlier.shape = NA)+facet_wrap(~group)+geom_jitter(width = 0.2)+stat_compare_means()+ theme_bw()+ theme(text = element_text(size = 20), axis.text = element_text(colour = "black"))+ ylab("scaled importance score")

left_join(mydata_ALL, paralogsDDR) %>% replace_na(list(paralogs = "FALSE")) %>% ggplot(aes(paralogs, scaled))+ geom_boxplot(outlier.shape = NA)+facet_wrap(~group)+geom_jitter(width = 0.2)+stat_compare_means()+ theme_bw()+ theme(text = element_text(size = 20), axis.text = element_text(colour = "black"))+ ylab("scaled importance score")

sbenfatto/PARIS documentation built on March 21, 2021, 4:39 a.m.