From 111909914c38b11eca066189587e725b351eac95 Mon Sep 17 00:00:00 2001
From: Bryan Roessler <bryanroessler@gmail.com>
Date: Mon, 2 Sep 2024 16:59:53 -0400
Subject: [PATCH] Before calculate_interaction_scores() refactor

---
 .../apps/r/calculate_interaction_zscores5.R   | 56 +++++++++++--------
 1 file changed, 32 insertions(+), 24 deletions(-)

diff --git a/workflow/apps/r/calculate_interaction_zscores5.R b/workflow/apps/r/calculate_interaction_zscores5.R
index d2fd120e..5cfefd9f 100644
--- a/workflow/apps/r/calculate_interaction_zscores5.R
+++ b/workflow/apps/r/calculate_interaction_zscores5.R
@@ -5,10 +5,17 @@ suppressMessages({
   library(dplyr)
   library(ggthemes)
   library(data.table)
+  library(unix)
 })
 
 options(warn = 2, max.print = 1000)
 
+
+# Set the memory limit to 30GB (30 * 1024 * 1024 * 1024 bytes)
+soft_limit <- 30 * 1024 * 1024 * 1024
+hard_limit <- 30 * 1024 * 1024 * 1024
+rlimit_as(soft_limit, hard_limit)
+
 # Constants for configuration
 plot_width <- 14
 plot_height <- 9
@@ -224,23 +231,6 @@ generate_and_save_plots <- function(df, output_dir, prefix, variables, include_q
   save_plots(prefix, plots, output_dir)
 }
 
-# Calculate summary statistics for all variables
-calculate_summary_stats <- function(df, variables, group_vars = c("conc_num", "conc_num_factor")) {
-  # Calculate summary statistics with the grouping columns
-  summary_stats <- df %>%
-    group_by(across(all_of(group_vars))) %>%
-    summarise(across(all_of(variables), list(
-      mean = ~mean(.x, na.rm = TRUE),
-      median = ~median(.x, na.rm = TRUE),
-      max = ~max(.x, na.rm = TRUE),
-      min = ~min(.x, na.rm = TRUE),
-      sd = ~sd(.x, na.rm = TRUE),
-      se = ~sd(.x, na.rm = TRUE) / sqrt(n() - 1)
-    ), .names = "{.col}_{.fn}"))
-  
-  return(summary_stats)
-}
-
 # Ensure all plots are saved and printed to PDF
 save_plots <- function(file_name, plot_list, output_dir) {
   # Save to PDF
@@ -294,17 +284,34 @@ process_strains <- function(df) {
   return(df_strains)
 }
 
+# Calculate summary statistics for all variables
+calculate_summary_stats <- function(df, variables, group_vars = c("conc_num", "conc_num_factor")) {
+  # Calculate summary statistics with the grouping columns
+  summary_stats <- df %>%
+    group_by(across(all_of(group_vars))) %>%
+    summarise(across(all_of(variables), list(
+      mean = ~mean(.x, na.rm = TRUE),
+      median = ~median(.x, na.rm = TRUE),
+      max = ~max(.x, na.rm = TRUE),
+      min = ~min(.x, na.rm = TRUE),
+      sd = ~sd(.x, na.rm = TRUE),
+      se = ~sd(.x, na.rm = TRUE) / sqrt(n() - 1)
+    ), .names = "{.col}_{.fn}"))
+  
+  return(summary_stats)
+}
 
-calculate_interaction_scores <- function(df, max_conc, variables, group_vars = c("OrfRep", "Gene", "num")) {
+calculate_interaction_scores <- function(df_ref, df, max_conc, variables, group_vars = c("OrfRep", "Gene", "num")) {
 
   # Pull the background means
   print("Calculating background means")
-  l_mean_bg <- df %>% filter(conc_num_factor == 0) %>% pull(L_mean)
-  k_mean_bg <- df %>% filter(conc_num_factor == 0) %>% pull(K_mean)
-  r_mean_bg <- df %>% filter(conc_num_factor == 0) %>% pull(r_mean)
-  auc_mean_bg <- df %>% filter(conc_num_factor == 0) %>% pull(AUC_mean)
+  l_mean_bg <- df_ref %>% filter(conc_num_factor == 0) %>% pull(L_mean)
+  k_mean_bg <- df_ref %>% filter(conc_num_factor == 0) %>% pull(K_mean)
+  r_mean_bg <- df_ref %>% filter(conc_num_factor == 0) %>% pull(r_mean)
+  auc_mean_bg <- df_ref %>% filter(conc_num_factor == 0) %>% pull(AUC_mean)
   
   # Calculate all necessary statistics and shifts in one step
+  print("Calculating interaction scores part 1")
   interaction_scores_all <- df %>%
     group_by(across(all_of(group_vars)), conc_num, conc_num_factor) %>%
     summarise(
@@ -344,6 +351,7 @@ calculate_interaction_scores <- function(df, max_conc, variables, group_vars = c
     ungroup()
   
   # Calculate linear models and interaction scores
+  print("Calculating interaction scores part 2")
   interaction_scores <- interaction_scores_all %>%
     group_by(across(all_of(group_vars))) %>%
     summarise(
@@ -648,9 +656,9 @@ main <- function() {
       # Calculate interactions
       variables <- c("L", "K", "r", "AUC")
       message("Calculating reference interaction scores")
-      reference_results <- calculate_interaction_scores(reference_strain, max_conc, variables)
+      reference_results <- calculate_interaction_scores(stats_joined, reference_strain, max_conc, variables)
       message("Calculating deletion interaction scores")
-      deletion_results <- calculate_interaction_scores(deletion_strains, max_conc, variables)
+      deletion_results <- calculate_interaction_scores(stats_joined, deletion_strains, max_conc, variables)
       
       zscores_calculations_reference <- reference_results$zscores_calculations
       zscores_interactions_reference <- reference_results$zscores_interactions