Example.Rmd

---
title: "Example"
output: html_document
date: "2024-05-10"
editor_options: 
  chunk_output_type: console
---

```{r setup, include=FALSE}
knitr::opts_chunk$set(echo = TRUE)
```

# Install and load packages and set up environment.
```{r}
options(scipen=999)
# devtools::install()

if (!requireNamespace("BiocManager", quietly = TRUE))
    {install.packages("BiocManager")}

required_packages <- c("limma", "fgsea", "Biobase")
for (k in required_packages){
  if (!requireNamespace(k, quietly = TRUE)) BiocManager::install(k)
  suppressPackageStartupMessages(library(k,character.only = TRUE))
 }
 


```
# Load data
```{r}
CADREmodel::load_and_preprocess_data	(drug_data_path= system.file("data", "drug_api.csv", package = "CADREmodel"),
                         cell_data_path= system.file("data", "cellline_api.csv", package = "CADREmodel"),
                         summary_df_ALMANAC_path=system.file("data", "summary_df_ALMANAC.csv", package = "CADREmodel"),
                         summary_df_api_path=system.file("data", "summary_api.csv", package = "CADREmodel"),
                         pathway_gene_set_path=system.file("data", "c2.cp.kegg.v2023.1.Hs.symbols.gmt", package = "CADREmodel"),
                         tissue_cancer_infosheet_path=system.file("data", "ALMANAC_tissues_GSE116450_cancer.csv", package = "CADREmodel"))


```
# Create folder and process data and prepare for variables
```{r}
# example here used the 9st row of sheet, aka melanoma.
row_number <- 9
tissue_cancer_infosheet_temp <- tissue_cancer_infosheet[row_number, ]
tissue <- tissue_cancer_infosheet_temp[1, 1]
cancer <- tissue_cancer_infosheet_temp[1, 2]
tissue_folder <- paste0("./result/", tissue, "/")
  
ALMANAC_sub_tissue_converted=CADREmodel::processTissueCancerInfo(tissue_cancer_infosheet, summary_df_ALMANAC, summary_df_api, DC_drug, DC_cell, row_number)

unique_cellnames <- unique(ALMANAC_sub_tissue_converted$cell_line_id)
unique_drugnames <- unique(c(ALMANAC_sub_tissue_converted$drug_row_id, ALMANAC_sub_tissue_converted$drug_col_id))

```
# Classify sensitivity to drug, and calculate Tri-scpre
```{r}

CADREmodel::analyzeDrugCellInteractions(ALMANAC_sub_tissue_converted, tissue_folder,unique_cellnames,unique_drugnames) 

# Extract results of this function
synergy_results <- readRDS(paste0(tissue_folder, "RDS/synergy_results.RDS"))
synergy_scores <- synergy_results$score_sheet
share_synergy_list <- synergy_results$share_synergy_list
share_antagonism_list <- synergy_results$share_antagonism_list

tri_scores=readRDS(paste0(tissue_folder,"RDS/tri_scores.RDS"))
sensitivity_classification=readRDS(paste0(tissue_folder,"RDS/sensitivity_classification.RDS"))

```
# Dysregulated Pathways
```{r}
geoID<- 'GSE116450'

interesting_pathways <- CADREmodel::analyzeDysregulatedPathways(geoID,sensitivity_classification,
                                                 pathway_gene_set,unique_drugnames,tissue_folder,tissue,cancer)

```
# Form graph
```{r}
graph_pathway=CADREmodel::generateGraph(interesting_pathways, tissue_folder, tissue) 

```
# Predict drug synergy from cluster
```{r}
probability_list_all=CADREmodel::prediction_by_cluster(tissue_folder, threshold_graph = 95,tri_scores,graph_pathway) 
```
# Calculate AUC and return prediction results
```{r}
drug_prediction=CADREmodel::calculate_performance_metrics(tissue_folder, tissue, threshold=0.99)
  
```