server-scripts/qhtcp-workflow/apps/r/calculate_interaction_zscores.R

suppressMessages({
  library("ggplot2")
  library("plotly")
  library("htmlwidgets")
  library("dplyr")
  library("rlang")
  library("ggthemes")
  library("data.table")
  library("unix")
})

options(warn = 2)

# Set the memory limit to 30GB (30 * 1024 * 1024 * 1024 bytes)
soft_limit <- 30 * 1024 * 1024 * 1024
hard_limit <- 30 * 1024 * 1024 * 1024
invisible(rlimit_as(soft_limit, hard_limit))

# Constants for configuration
plot_width <- 14
plot_height <- 9
base_size <- 14

parse_arguments <- function() {
  args <- if (interactive()) {
    c(
      "/home/bryan/documents/develop/hartmanlab/qhtcp-workflow/out/20240116_jhartman2_DoxoHLD/20240116_jhartman2_DoxoHLD",
      "/home/bryan/documents/develop/hartmanlab/qhtcp-workflow/apps/r/SGD_features.tab",
      "/home/bryan/documents/develop/hartmanlab/qhtcp-workflow/out/20240116_jhartman2_DoxoHLD/easy/20240116_jhartman2_DoxoHLD/results_std.txt",
      "/home/bryan/documents/develop/hartmanlab/qhtcp-workflow/out/20240116_jhartman2_DoxoHLD/20240822_jhartman2_DoxoHLD/exp1",
      "Experiment 1: Doxo versus HLD",
      3,
      "/home/bryan/documents/develop/hartmanlab/qhtcp-workflow/out/20240116_jhartman2_DoxoHLD/20240822_jhartman2_DoxoHLD/exp2",
      "Experiment 2: HLD versus Doxo",
      3
    )
  } else {
    commandArgs(trailingOnly = TRUE)
  }

  out_dir <- normalizePath(args[1], mustWork = FALSE)
  sgd_gene_list <- normalizePath(args[2], mustWork = FALSE)
  easy_results_file <- normalizePath(args[3], mustWork = FALSE)

  # The remaining arguments should be in groups of 3
  exp_args <- args[-(1:3)]
  if (length(exp_args) %% 3 != 0) {
    stop("Experiment arguments should be in groups of 3: path, name, sd.")
  }

  experiments <- list()
  for (i in seq(1, length(exp_args), by = 3)) {
    exp_name <- exp_args[i + 1]
    experiments[[exp_name]] <- list(
      path = normalizePath(exp_args[i], mustWork = FALSE),
      sd = as.numeric(exp_args[i + 2])
    )
  }

  list(
    out_dir = out_dir,
    sgd_gene_list = sgd_gene_list,
    easy_results_file = easy_results_file,
    experiments = experiments
  )
}

args <- parse_arguments()

# Should we keep output in exp dirs or combine in the study output dir?
# dir.create(file.path(args$out_dir, "zscores"), showWarnings = FALSE)
# dir.create(file.path(args$out_dir, "zscores", "qc"), showWarnings = FALSE)

# Define themes and scales
theme_publication <- function(base_size = 14, base_family = "sans", legend_position = "bottom") {
  theme_foundation <- ggplot2::theme_grey(base_size = base_size, base_family = base_family)
  
  theme_foundation %+replace%
    theme(
      plot.title = element_text(face = "bold", size = rel(1.2), hjust = 0.5),
      text = element_text(),
      panel.background = element_rect(colour = NA),
      plot.background = element_rect(colour = NA),
      panel.border = element_rect(colour = NA),
      axis.title = element_text(face = "bold", size = rel(1)),
      axis.title.y = element_text(angle = 90, vjust = 2, size = 18),
      axis.title.x = element_text(vjust = -0.2, size = 18),
      axis.line = element_line(colour = "black"),
      axis.text.x = element_text(size = 16),
      axis.text.y = element_text(size = 16),
      panel.grid.major = element_line(colour = "#f0f0f0"),
      panel.grid.minor = element_blank(),
      legend.key = element_rect(colour = NA),
      legend.position = legend_position,
      legend.direction = ifelse(legend_position == "right", "vertical", "horizontal"),
      plot.margin = unit(c(10, 5, 5, 5), "mm"),
      strip.background = element_rect(colour = "#f0f0f0", fill = "#f0f0f0"),
      strip.text = element_text(face = "bold")
    )
}

scale_fill_publication <- function(...) {
  discrete_scale("fill", "Publication", manual_pal(values = c(
    "#386cb0", "#fdb462", "#7fc97f", "#ef3b2c", "#662506",
    "#a6cee3", "#fb9a99", "#984ea3", "#ffff33"
  )), ...)
}

scale_colour_publication <- function(...) {
  discrete_scale("colour", "Publication", manual_pal(values = c(
    "#386cb0", "#fdb462", "#7fc97f", "#ef3b2c", "#662506",
    "#a6cee3", "#fb9a99", "#984ea3", "#ffff33"
  )), ...)
}

# Load the initial dataframe from the easy_results_file
load_and_process_data <- function(easy_results_file, sd = 3) {
  df <- read.delim(easy_results_file, skip = 2, as.is = TRUE, row.names = 1, strip.white = TRUE)

  df <- df %>%
    filter(!(.[[1]] %in% c("", "Scan"))) %>%
    filter(!is.na(ORF) & ORF != "" & !Gene %in% c("BLANK", "Blank", "blank") & Drug != "BMH21") %>%
    # Rename columns
    rename(L = l, num = Num., AUC = AUC96, scan = Scan, last_bg = LstBackgrd, first_bg = X1stBackgrd) %>%
    mutate(
      across(c(Col, Row, num, L, K, r, scan, AUC, last_bg, first_bg), as.numeric),
      delta_bg = last_bg - first_bg,
      delta_bg_tolerance = mean(delta_bg, na.rm = TRUE) + (sd * sd(delta_bg, na.rm = TRUE)),
      NG = if_else(L == 0 & !is.na(L), 1, 0),
      DB = if_else(delta_bg >= delta_bg_tolerance, 1, 0),
      SM = 0,
      OrfRep = if_else(ORF == "YDL227C", "YDL227C", OrfRep), # should these be hardcoded?
      conc_num = as.numeric(gsub("[^0-9\\.]", "", Conc))
    ) %>%
    mutate(
      conc_num_factor = as.factor(match(conc_num, sort(unique(conc_num))) - 1)
    )
    
    return(df)
  }

# Update Gene names using the SGD gene list
update_gene_names <- function(df, sgd_gene_list) {
  # Load SGD gene list
  genes <- read.delim(file = sgd_gene_list,
    quote = "", header = FALSE,
    colClasses = c(rep("NULL", 3), rep("character", 2), rep("NULL", 11)))
  
  # Create a named vector for mapping ORF to GeneName
  gene_map <- setNames(genes$V5, genes$V4)
  # Vectorized match to find the GeneName from gene_map
  mapped_genes <- gene_map[df$ORF]
  # Replace NAs in mapped_genes with original Gene names (preserves existing Gene names if ORF is not found)
  updated_genes <- ifelse(is.na(mapped_genes) | df$OrfRep == "YDL227C", df$Gene, mapped_genes)
  # Ensure Gene is not left blank or incorrectly updated to "OCT1"
  df <- df %>%
    mutate(Gene = ifelse(updated_genes == "" | updated_genes == "OCT1", OrfRep, updated_genes))
  
  return(df)
}

calculate_summary_stats <- function(df, variables, group_vars) {

  summary_stats <- df %>%
    group_by(across(all_of(group_vars))) %>%
    summarise(
      N = n(),
      across(all_of(variables), list(
        mean = ~mean(., na.rm = TRUE),
        median = ~median(., na.rm = TRUE),
        max = ~ ifelse(all(is.na(.)), NA, max(., na.rm = TRUE)),
        min = ~ ifelse(all(is.na(.)), NA, min(., na.rm = TRUE)),
        sd = ~sd(., na.rm = TRUE),
        se = ~sd(., na.rm = TRUE) / sqrt(N) - 1 # TODO needs comment for explanation
      ), .names = "{.fn}_{.col}"),
      .groups = "drop"
    )

  # Create a cleaned version of df that doesn't overlap with summary_stats
  cols_to_keep <- setdiff(names(df), names(summary_stats)[-which(names(summary_stats) %in% group_vars)])
  df_cleaned <- df %>%
    select(all_of(cols_to_keep))

  df_with_stats <- left_join(df_cleaned, summary_stats, by = group_vars)

  return(list(summary_stats = summary_stats, df_with_stats = df_with_stats))
}

calculate_interaction_scores <- function(df, max_conc) {

  variables <- c("L", "K", "r", "AUC")
  group_vars <- c("OrfRep", "Gene", "num")

  # Calculate total concentration variables
  total_conc_num <- length(unique(df$conc_num))

  # Pull the background means and standard deviations from zero concentration
  bg_means <- list(
    L = df %>% filter(conc_num == 0) %>% pull(mean_L) %>% first(),
    K = df %>% filter(conc_num == 0) %>% pull(mean_K) %>% first(),
    r = df %>% filter(conc_num == 0) %>% pull(mean_r) %>% first(),
    AUC = df %>% filter(conc_num == 0) %>% pull(mean_AUC) %>% first()
  )

  bg_sd <- list(
    L = df %>% filter(conc_num == 0) %>% pull(sd_L) %>% first(),
    K = df %>% filter(conc_num == 0) %>% pull(sd_K) %>% first(),
    r = df %>% filter(conc_num == 0) %>% pull(sd_r) %>% first(),
    AUC = df %>% filter(conc_num == 0) %>% pull(sd_AUC) %>% first()
  )

  # Calculate per spot
  stats <- calculate_summary_stats(df,
    variables = variables,
    group_vars = c("OrfRep", "Gene", "num", "conc_num", "conc_num_factor")
    )$summary_stats

  stats <- df %>%
    group_by(across(all_of(group_vars))) %>%
    mutate(
      WT_L = mean_L,
      WT_K = mean_K,
      WT_r = mean_r,
      WT_AUC = mean_AUC,
      WT_sd_L = sd_L,
      WT_sd_K = sd_K,
      WT_sd_r = sd_r,
      WT_sd_AUC = sd_AUC
    )

  stats <- stats %>%
    mutate(
      Raw_Shift_L = first(mean_L) - bg_means$L,
      Raw_Shift_K = first(mean_K) - bg_means$K,
      Raw_Shift_r = first(mean_r) - bg_means$r,
      Raw_Shift_AUC = first(mean_AUC) - bg_means$AUC,
      Z_Shift_L = first(Raw_Shift_L) / bg_sd$L,
      Z_Shift_K = first(Raw_Shift_K) / bg_sd$K,
      Z_Shift_r = first(Raw_Shift_r) / bg_sd$r,
      Z_Shift_AUC = first(Raw_Shift_AUC) / bg_sd$AUC
    )

  stats <- stats %>%
    mutate(
      Exp_L = WT_L + Raw_Shift_L,
      Exp_K = WT_K + Raw_Shift_K,
      Exp_r = WT_r + Raw_Shift_r,
      Exp_AUC = WT_AUC + Raw_Shift_AUC,
      Delta_L = mean_L - Exp_L,
      Delta_K = mean_K - Exp_K,
      Delta_r = mean_r - Exp_r,
      Delta_AUC = mean_AUC - Exp_AUC
    )

  stats <- stats %>%
    mutate(
      Delta_L = if_else(NG == 1, mean_L - WT_L, Delta_L),
      Delta_K = if_else(NG == 1, mean_K - WT_K, Delta_K),
      Delta_r = if_else(NG == 1, mean_r - WT_r, Delta_r),
      Delta_AUC = if_else(NG == 1, mean_AUC - WT_AUC, Delta_AUC),
      Delta_L = if_else(SM == 1, mean_L - WT_L, Delta_L)
    )

  stats <- stats %>%
    mutate(
      Zscore_L = Delta_L / WT_sd_L,
      Zscore_K = Delta_K / WT_sd_K,
      Zscore_r = Delta_r / WT_sd_r,
      Zscore_AUC = Delta_AUC / WT_sd_AUC
    )

  # Calculate linear models
  lm_L <- lm(Delta_L ~ conc_num, data = stats)
  lm_K <- lm(Delta_K ~ conc_num, data = stats)
  lm_r <- lm(Delta_r ~ conc_num, data = stats)
  lm_AUC <- lm(Delta_AUC ~ conc_num, data = stats)

  interactions <- stats %>%
    group_by(across(all_of(group_vars))) %>%
    summarise(
      OrfRep = first(OrfRep),
      Gene = first(Gene),
      num = first(num),
      conc_num = first(conc_num),
      conc_num_factor = first(conc_num_factor),
      Raw_Shift_L = first(Raw_Shift_L),
      Raw_Shift_K = first(Raw_Shift_K),
      Raw_Shift_r = first(Raw_Shift_r),
      Raw_Shift_AUC = first(Raw_Shift_AUC),
      Z_Shift_L = first(Z_Shift_L),
      Z_Shift_K = first(Z_Shift_K),
      Z_Shift_r = first(Z_Shift_r),
      Z_Shift_AUC = first(Z_Shift_AUC),
      Sum_Zscore_L = sum(Zscore_L, na.rm = TRUE),
      Sum_Zscore_K = sum(Zscore_K, na.rm = TRUE),
      Sum_Zscore_r = sum(Zscore_r, na.rm = TRUE),
      Sum_Zscore_AUC = sum(Zscore_AUC, na.rm = TRUE),
      lm_Score_L = max_conc * coef(lm_L)[2] + coef(lm_L)[1],
      lm_Score_K = max_conc * coef(lm_K)[2] + coef(lm_K)[1],
      lm_Score_r = max_conc * coef(lm_r)[2] + coef(lm_r)[1],
      lm_Score_AUC = max_conc * coef(lm_AUC)[2] + coef(lm_AUC)[1],
      R_Squared_L = summary(lm_L)$r.squared,
      R_Squared_K = summary(lm_K)$r.squared,
      R_Squared_r = summary(lm_r)$r.squared,
      R_Squared_AUC = summary(lm_AUC)$r.squared,
      lm_intercept_L = coef(lm_L)[1],
      lm_slope_L = coef(lm_L)[2],
      lm_intercept_K = coef(lm_K)[1],
      lm_slope_K = coef(lm_K)[2],
      lm_intercept_r = coef(lm_r)[1],
      lm_slope_r = coef(lm_r)[2],
      lm_intercept_AUC = coef(lm_AUC)[1],
      lm_slope_AUC = coef(lm_AUC)[2],
      NG = sum(NG, na.rm = TRUE),
      DB = sum(DB, na.rm = TRUE),
      SM = sum(SM, na.rm = TRUE),
      .groups = "keep"
    )

  num_non_removed_concs <- total_conc_num - sum(stats$DB, na.rm = TRUE) - 1
  
  interactions <- interactions %>%
    mutate(
      Avg_Zscore_L = Sum_Zscore_L / num_non_removed_concs,
      Avg_Zscore_K = Sum_Zscore_K / num_non_removed_concs,
      Avg_Zscore_r = Sum_Zscore_r / (total_conc_num - 1),
      Avg_Zscore_AUC = Sum_Zscore_AUC / (total_conc_num - 1),
      Z_lm_L = (lm_Score_L - mean(lm_Score_L, na.rm = TRUE)) / sd(lm_Score_L, na.rm = TRUE),
      Z_lm_K = (lm_Score_K - mean(lm_Score_K, na.rm = TRUE)) / sd(lm_Score_K, na.rm = TRUE),
      Z_lm_r = (lm_Score_r - mean(lm_Score_r, na.rm = TRUE)) / sd(lm_Score_r, na.rm = TRUE),
      Z_lm_AUC = (lm_Score_AUC - mean(lm_Score_AUC, na.rm = TRUE)) / sd(lm_Score_AUC, na.rm = TRUE)
    ) %>%
    arrange(desc(Z_lm_L), desc(NG))

  # Declare column order for output
  calculations <- stats %>%
    select(
      "OrfRep", "Gene", "num", "conc_num", "conc_num_factor", "N",
      "mean_L", "mean_K", "mean_r", "mean_AUC",
      "median_L", "median_K", "median_r", "median_AUC",
      "sd_L", "sd_K", "sd_r", "sd_AUC",
      "se_L", "se_K", "se_r", "se_AUC",
      "Raw_Shift_L", "Raw_Shift_K", "Raw_Shift_r", "Raw_Shift_AUC",
      "Z_Shift_L", "Z_Shift_K", "Z_Shift_r", "Z_Shift_AUC",
      "WT_L", "WT_K", "WT_r", "WT_AUC",
      "WT_sd_L", "WT_sd_K", "WT_sd_r", "WT_sd_AUC",
      "Exp_L", "Exp_K", "Exp_r", "Exp_AUC",
      "Delta_L", "Delta_K", "Delta_r", "Delta_AUC",
      "Zscore_L", "Zscore_K", "Zscore_r", "Zscore_AUC",
      "NG", "SM", "DB")
    
  calculations_joined <- df %>% select(-any_of(setdiff(names(calculations), c("OrfRep", "Gene", "num", "conc_num", "conc_num_factor"))))
  calculations_joined <- left_join(calculations_joined, calculations, by = c("OrfRep", "Gene", "num", "conc_num", "conc_num_factor"))

  interactions_joined <- df %>% select(-any_of(setdiff(names(interactions), c("OrfRep", "Gene", "num", "conc_num", "conc_num_factor"))))
  interactions_joined <- left_join(interactions_joined, interactions, by = c("OrfRep", "Gene", "num", "conc_num", "conc_num_factor"))

  return(list(calculations = calculations, interactions = interactions, interactions_joined = interactions_joined,
    calculations_joined = calculations_joined))
}

generate_and_save_plots <- function(out_dir, file_name, plot_configs, grid_layout = NULL) {
  message("Generating ", file_name, ".pdf and ", file_name, ".html")

  # Prepare lists to collect plots
  static_plots <- list()
  plotly_plots <- list()

  for (i in seq_along(plot_configs)) {
    config <- plot_configs[[i]]
    df <- config$df

    aes_mapping <- if (is.null(config$color_var)) {
      if (is.null(config$y_var)) {
        aes(x = .data[[config$x_var]])
      } else {
        aes(x = .data[[config$x_var]], y = .data[[config$y_var]])
      }
    } else {
      if (is.null(config$y_var)) {
        aes(x = .data[[config$x_var]], color = as.factor(.data[[config$color_var]]))
      } else {
        aes(x = .data[[config$x_var]], y = .data[[config$y_var]], color = as.factor(.data[[config$color_var]]))
      }
    }

    # Start building the plot with aes_mapping
    plot_base <- ggplot(df, aes_mapping)

    # Use appropriate helper function based on plot type
    plot <- switch(config$plot_type,
      "scatter" = generate_scatter_plot(plot_base, config),
      "box" = generate_box_plot(plot_base, config),
      "density" = plot_base + geom_density(),
      "bar" = plot_base + geom_bar(),
      plot_base  # default case if no type matches
    )

    # Apply additional settings
    if (!is.null(config$legend_position)) {
      plot <- plot + theme(legend.position = config$legend_position)
    }

    # Add title and labels
    if (!is.null(config$title)) {
      plot <- plot + ggtitle(config$title)
    }
    if (!is.null(config$x_label)) {
      plot <- plot + xlab(config$x_label)
    }
    if (!is.null(config$y_label)) {
      plot <- plot + ylab(config$y_label)
    }

    # Apply scale_color_discrete(guide = FALSE) when color_var is NULL
    if (is.null(config$color_var)) {
      plot <- plot + scale_color_discrete(guide = "none")
    }

    # Add interactive tooltips for plotly
    tooltip_vars <- c()
    if (config$plot_type == "scatter") {
      if (!is.null(config$delta_bg_point) && config$delta_bg_point) {
        tooltip_vars <- c(tooltip_vars, "OrfRep", "Gene", "delta_bg")
      } else if (!is.null(config$gene_point) && config$gene_point) {
        tooltip_vars <- c(tooltip_vars, "OrfRep", "Gene")
      } else if (!is.null(config$y_var) && !is.null(config$x_var)) {
        tooltip_vars <- c(config$x_var, config$y_var)
      }
    }

    # Convert to plotly object
    if (length(tooltip_vars) > 0) {
      plotly_plot <- ggplotly(plot, tooltip = tooltip_vars)
    } else {
      plotly_plot <- ggplotly(plot, tooltip = "none")
    }

    # Adjust legend position if specified
    if (!is.null(config$legend_position) && config$legend_position == "bottom") {
      plotly_plot <- plotly_plot %>% layout(legend = list(orientation = "h"))
    }

    # Add plots to lists
    static_plots[[i]] <- plot
    plotly_plots[[i]] <- plotly_plot
  }

  # Save static PDF plots
  pdf(file.path(out_dir, paste0(file_name, ".pdf")), width = 14, height = 9)
  lapply(static_plots, print)
  dev.off()

  # Combine and save interactive HTML plots
  combined_plot <- subplot(
    plotly_plots,
    nrows = if (!is.null(grid_layout) && !is.null(grid_layout$nrow)) {
      grid_layout$nrow
    } else {
      # Calculate nrow based on the length of plotly_plots (default 1 row if only one plot)
      ceiling(length(plotly_plots) / ifelse(!is.null(grid_layout) && !is.null(grid_layout$ncol), grid_layout$ncol, 1))
    },
    margin = 0.05
  )
  saveWidget(combined_plot, file = file.path(out_dir, paste0(file_name, ".html")), selfcontained = TRUE)
}

generate_scatter_plot <- function(plot, config) {
  
  # Determine Shape, Size, and Position for geom_point
  shape <- if (!is.null(config$shape)) config$shape else 3
  size <- if (!is.null(config$size)) config$size else 0.1
  position <- if (!is.null(config$position) && config$position == "jitter") "jitter" else "identity"
  
  # Add geom_point with determined parameters
  plot <- plot + geom_point(
    shape = shape,
    size = size,
    position = position
  )

  if (!is.null(config$cyan_points)) {
    plot <- plot + geom_point(
      aes(x = .data[[config$x_var]], y = .data[[config$y_var]]),
      color = "cyan",
      shape = 3,
      size = 0.5
    )
  }
  
  # Add Smooth Line if specified
  if (!is.null(config$add_smooth) && config$add_smooth) {
    if (!is.null(config$lm_line)) {
      plot <- plot +
        geom_abline(
          intercept = config$lm_line$intercept,
          slope = config$lm_line$slope,
          color = "blue"
        )
    } else {
      plot <- plot +
        geom_smooth(
          method = "lm",
          se = FALSE,
          color = "blue"
        )
    }
  }
  
  # Add SD Bands if specified
  if (!is.null(config$sd_band_values)) {
    for (sd_band in config$sd_band_values) {
      plot <- plot +
        annotate(
          "rect",
          xmin = -Inf, xmax = Inf,
          ymin = sd_band, ymax = Inf,
          fill = "#542788",
          alpha = 0.3
        ) +
        annotate(
          "rect",
          xmin = -Inf, xmax = Inf,
          ymin = -sd_band, ymax = -Inf,
          fill = "orange",
          alpha = 0.3
        ) +
        geom_hline(
          yintercept = c(-sd_band, sd_band),
          color = "gray"
        )
    }
  }

  # Add Rectangles if specified
  if (!is.null(config$rectangles)) {
    for (rect in config$rectangles) {
      plot <- plot + annotate(
        "rect",
        xmin = rect$xmin,
        xmax = rect$xmax,
        ymin = rect$ymin,
        ymax = rect$ymax,
        fill = ifelse(is.null(rect$fill), NA, rect$fill),
        color = ifelse(is.null(rect$color), "black", rect$color),
        alpha = ifelse(is.null(rect$alpha), 0.1, rect$alpha)
      )
    }
  }
  
  # Add Error Bars if specified
  if (!is.null(config$error_bar) && config$error_bar && !is.null(config$y_var)) {
    y_mean_col <- paste0("mean_", config$y_var)
    y_sd_col <- paste0("sd_", config$y_var)
    
    plot <- plot +
      geom_errorbar(
        aes(
          ymin = !!sym(y_mean_col) - !!sym(y_sd_col),
          ymax = !!sym(y_mean_col) + !!sym(y_sd_col)
        ),
        alpha = 0.3
      )
  }
  
  # Customize X-axis if specified
  if (!is.null(config$x_breaks) && !is.null(config$x_labels) && !is.null(config$x_label)) {
    plot <- plot +
      scale_x_discrete(
        name = config$x_label,
        breaks = config$x_breaks,
        labels = config$x_labels
      )
  }
  
  # Apply coord_cartesian if specified
  if (!is.null(config$coord_cartesian)) {
    plot <- plot + coord_cartesian(ylim = config$coord_cartesian)
  }
  
  # Set Y-axis limits if specified
  if (!is.null(config$ylim_vals)) {
    plot <- plot + scale_y_continuous(limits = config$ylim_vals)
  }
  
  # Add Annotations if specified
  if (!is.null(config$annotations)) {
    for (annotation in config$annotations) {
      plot <- plot +
        annotate(
          "text",
          x = annotation$x,
          y = annotation$y,
          label = annotation$label,
          na.rm = TRUE
        )
    }
  }
  
  # Add Title if specified
  if (!is.null(config$title)) {
    plot <- plot + ggtitle(config$title)
  }
  
  # Adjust Legend Position if specified
  if (!is.null(config$legend_position)) {
    plot <- plot + theme(legend.position = config$legend_position)
  }
  
  return(plot)
}

generate_box_plot <- function(plot, config) {
  plot <- plot + geom_boxplot()
  
  if (!is.null(config$x_breaks) && !is.null(config$x_labels) && !is.null(config$x_label)) {
    plot <- plot + scale_x_discrete(
      name = config$x_label,
      breaks = config$x_breaks,
      labels = config$x_labels
    )
  }

  if (!is.null(config$coord_cartesian)) {
    plot <- plot + coord_cartesian(ylim = config$coord_cartesian)
  }

  return(plot)
}

generate_plate_analysis_plot_configs <- function(variables, stages = c("before", "after"),
  df_before = NULL, df_after = NULL, plot_type = "scatter") {
  plots <- list()
  for (var in variables) {
    for (stage in stages) {
      df_plot <- if (stage == "before") df_before else df_after
      
      # Adjust settings based on plot_type
      if (plot_type == "scatter") {
        error_bar <- TRUE
        position <- "jitter"
      } else if (plot_type == "box") {
        error_bar <- FALSE
        position <- NULL
      }
      
      config <- list(
        df = df_plot,
        x_var = "scan",
        y_var = var,
        plot_type = plot_type,
        title = paste("Plate analysis by Drug Conc for", var, stage, "quality control"),
        error_bar = error_bar,
        color_var = "conc_num_factor",
        position = position
      )
      plots <- append(plots, list(config))
    }
  }
  return(plots)
}

generate_interaction_plot_configs <- function(df, variables) {

  configs <- list()

  limits_map <- list(
    L = c(-65, 65),
    K = c(-65, 65),
    r = c(-0.65, 0.65),
    AUC = c(-6500, 6500)
  )

  df_filtered <- filter_data(df, variables, missing = TRUE, limits_map = limits_map)

  # Define annotation label functions
  generate_annotation_labels <- function(df, var, annotation_name) {
    switch(annotation_name,
      ZShift = paste("ZShift =", round(df[[paste0("Z_Shift_", var)]], 2)),
      lm_ZScore = paste("lm ZScore =", round(df[[paste0("Z_lm_", var)]], 2)),
      NG = paste("NG =", df$NG),
      DB = paste("DB =", df$DB),
      SM = paste("SM =", df$SM),
      NULL  # Default case for unrecognized annotation names
    )
  }

  # Define annotation positions relative to the y-axis range
  calculate_annotation_positions <- function(y_range) {
    y_min <- min(y_range)
    y_max <- max(y_range)
    y_span <- y_max - y_min

    list(
      ZShift = y_max - 0.1 * y_span,
      lm_ZScore = y_max - 0.2 * y_span,
      NG = y_min + 0.2 * y_span,
      DB = y_min + 0.1 * y_span,
      SM = y_min + 0.05 * y_span
    )
  }

  # Create configurations for each variable
  for (variable in variables) {

    y_range <- limits_map[[variable]]
    annotation_positions <- calculate_annotation_positions(y_range)
    lm_line <- list(
      intercept = df_filtered[[paste0("lm_intercept_", variable)]],
      slope = df_filtered[[paste0("lm_slope_", variable)]]
    )

    # Determine x-axis midpoint
    num_levels <- length(levels(df_filtered$conc_num_factor))
    x_pos <- (1 + num_levels) / 2  # Midpoint of x-axis

    # Generate annotations
    annotations <- lapply(names(annotation_positions), function(annotation_name) {
      label <- generate_annotation_labels(df_filtered, variable, annotation_name)
      y_pos <- annotation_positions[[annotation_name]]
      if (!is.null(label)) {
        list(x = x_pos, y = y_pos, label = label)
      } else {
        message(paste("Warning: No annotation found for", annotation_name))
        NULL
      }
    })

    # Remove NULL annotations
    annotations <- Filter(Negate(is.null), annotations)

    # Shared plot settings
    plot_settings <- list(
      df = df_filtered,
      x_var = "conc_num_factor",
      y_var = variable,
      ylim_vals = y_range,
      annotations = annotations,
      lm_line = lm_line,
      x_breaks = levels(df_filtered$conc_num_factor),
      x_labels = levels(df_filtered$conc_num_factor),
      x_label = unique(df_filtered$Drug[1]),
      coord_cartesian = y_range  # Use the actual y-limits
    )

    # Scatter plot config
    configs[[length(configs) + 1]] <- modifyList(plot_settings, list(
      plot_type = "scatter",
      title = sprintf("%s      %s", df_filtered$OrfRep[1], df_filtered$Gene[1]),
      error_bar = TRUE,
      position = "jitter"
    ))

    # Box plot config
    configs[[length(configs) + 1]] <- modifyList(plot_settings, list(
      plot_type = "box",
      title = sprintf("%s      %s (box plot)", df_filtered$OrfRep[1], df_filtered$Gene[1]),
      error_bar = FALSE
    ))
  }

  return(configs)
}

generate_rank_plot_configs <- function(df_filtered, variables, is_lm = FALSE) {
  
  sd_bands <- c(1, 2, 3)
  
  configs <- list()
  
  # SD-based plots for L and K
  for (variable in c("L", "K")) {
    for (sd_band in sd_bands) {
      # Determine columns based on whether it's lm or not
      if (is_lm) {
        rank_var <- paste0("Rank_lm_", variable)
        zscore_var <- paste0("Z_lm_", variable)
        y_label <- paste("Int Z score", variable)
      } else {
        rank_var <- paste0("Rank_", variable)
        zscore_var <- paste0("Avg_Zscore_", variable)
        y_label <- paste("Avg Z score", variable)
      }
      
      num_enhancers <- sum(df_filtered[[zscore_var]] >= sd_band, na.rm = TRUE)
      num_suppressors <- sum(df_filtered[[zscore_var]] <= -sd_band, na.rm = TRUE)
      
      # Annotated plot configuration
      configs[[length(configs) + 1]] <- list(
        df = df_filtered,
        x_var = rank_var,
        y_var = zscore_var,
        plot_type = "scatter",
        title = paste(y_label, "vs. Rank for", variable, "above", sd_band, "SD"),
        sd_band = sd_band,
        annotations = list(
          list(
            x = median(df_filtered[[rank_var]], na.rm = TRUE),
            y = 10,
            label = paste("Deletion Enhancers =", num_enhancers)
          ),
          list(
            x = median(df_filtered[[rank_var]], na.rm = TRUE),
            y = -10,
            label = paste("Deletion Suppressors =", num_suppressors)
          )
        ),
        sd_band_values = sd_band,
        shape = 3,
        size = 0.1,
        y_label = y_label,
        x_label = "Rank",
        legend_position = "none"
      )
      
      # Non-Annotated Plot Configuration
      configs[[length(configs) + 1]] <- list(
        df = df_filtered,
        x_var = rank_var,
        y_var = zscore_var,
        plot_type = "scatter",
        title = paste(y_label, "vs. Rank for", variable, "above", sd_band, "SD No Annotations"),
        sd_band = sd_band,
        annotations = NULL,
        sd_band_values = sd_band,
        shape = 3,
        size = 0.1,
        y_label = y_label,
        x_label = "Rank",
        legend_position = "none"
      )
    }
  }
    
  # Avg ZScore and Rank Avg ZScore Plots for r, L, K, and AUC
  for (variable in variables) {
    for (plot_type in c("Avg_Zscore_vs_lm", "Rank_Avg_Zscore_vs_lm")) {
      # Define specific variables based on plot type
      if (plot_type == "Avg_Zscore_vs_lm") {
        x_var <- paste0("Avg_Zscore_", variable)
        y_var <- paste0("Z_lm_", variable)
        title_suffix <- paste("Avg Zscore vs lm", variable)
        rectangles <- list(
          list(xmin = -2, xmax = 2, ymin = -2, ymax = 2,
            fill = NA, color = "grey20", alpha = 0.1
          )
        )
      } else {
        x_var <- paste0("Rank_", variable)
        y_var <- paste0("Rank_lm_", variable)
        title_suffix <- paste("Rank Avg Zscore vs lm", variable)
        rectangles <- NULL
      }
      
      # Fit linear model
      lm_model <- lm(as.formula(paste(y_var, "~", x_var)), data = df_filtered)
      lm_summary <- summary(lm_model)
      
      # Extract intercept and slope from the model coefficients
      intercept <- coef(lm_model)[1]
      slope <- coef(lm_model)[2]
        
      configs[[length(configs) + 1]] <- list(
        df = df_filtered,
        x_var = x_var,
        y_var = y_var,
        plot_type = "scatter",
        title = title_suffix,
        annotations = list(
          list(
            x = 0,
            y = 0,
            label = paste("R-squared =", round(lm_summary$r.squared, 2))
          )
        ),
        sd_band_values = NULL, # Not applicable
        shape = 3,
        size = 0.1,
        add_smooth = TRUE,
        lm_line = list(intercept = intercept, slope = slope),
        legend_position = "right",
        color_var = "Overlap",
        x_label = x_var,
        y_label = y_var,
        rectangles = rectangles
      )
    }
  }
  return(configs)
}

generate_correlation_plot_configs <- function(df) {
  # Define relationships for plotting
  relationships <- list(
    list(x = "Z_lm_L", y = "Z_lm_K", label = "Interaction L vs. Interaction K"),
    list(x = "Z_lm_L", y = "Z_lm_r", label = "Interaction L vs. Interaction r"),
    list(x = "Z_lm_L", y = "Z_lm_AUC", label = "Interaction L vs. Interaction AUC"),
    list(x = "Z_lm_K", y = "Z_lm_r", label = "Interaction K vs. Interaction r"),
    list(x = "Z_lm_K", y = "Z_lm_AUC", label = "Interaction K vs. Interaction AUC"),
    list(x = "Z_lm_r", y = "Z_lm_AUC", label = "Interaction r vs. Interaction AUC")
  )

  configs <- list()

  for (rel in relationships) {
    # Fit linear model
    lm_model <- lm(as.formula(paste(rel$y, "~", rel$x)), data = df)
    lm_summary <- summary(lm_model)

    # Construct plot configuration
    config <- list(
      df = df,
      x_var = rel$x,
      y_var = rel$y,
      plot_type = "scatter",
      title = rel$label,
      x_label = paste("z-score", gsub("Z_lm_", "", rel$x)),
      y_label = paste("z-score", gsub("Z_lm_", "", rel$y)),
      annotations = list(
        list(x = 0, y = 0, label = paste("R-squared =", round(lm_summary$r.squared, 3)))
      ),
      add_smooth = TRUE,  # Add regression line
      lm_line = list(intercept = coef(lm_model)[1], slope = coef(lm_model)[2]),
      legend_position = "right",
      shape = 3,
      size = 0.5,
      color_var = "Overlap",
      rectangles = list(
        list(
          xmin = -2, xmax = 2, ymin = -2, ymax = 2,
          fill = NA, color = "grey20", alpha = 0.1
        )
      ),
      cyan_points = TRUE
    )

    configs[[length(configs) + 1]] <- config
  }

  return(configs)
}

filter_data <- function(df, variables, nf = FALSE, missing = FALSE, adjust = FALSE,
  rank = FALSE, limits_map = NULL, verbose = TRUE) {

  avg_zscore_cols <- paste0("Avg_Zscore_", variables)
  z_lm_cols <- paste0("Z_lm_", variables)

  # Step 1: Adjust NAs to 0.001 for linear model (if adjust = TRUE)
  if (adjust) {
    if (verbose) message("Replacing NA with 0.001 for Avg_Zscore_ and Z_lm_ columns")
    df <- df %>%
      mutate(
        across(all_of(avg_zscore_cols), ~ ifelse(is.na(.), 0.001, .)),
        across(all_of(z_lm_cols), ~ ifelse(is.na(.), 0.001, .))
      )
  }

  # Filter non-finite values
  if (nf) {
    df <- df %>%
      filter(if_all(all_of(variables), ~ is.finite(.)))
  }

  # Filter missing values
  if (missing) {
    df <- df %>%
      filter(if_all(all_of(variables), ~ !is.na(.)))
  }

  # Apply limits from 'limits_map' if provided
  if (!is.null(limits_map)) {
    for (variable in names(limits_map)) {
      if (variable %in% variables) {
        ylim_vals <- limits_map[[variable]]
        df <- df %>%
          filter(.data[[variable]] >= ylim_vals[1] & .data[[variable]] <= ylim_vals[2])
      }
    }
  }

  # Calculate and add rank columns
  if (rank) {
    if (verbose) message("Calculating ranks for variable(s): ", paste(variables, collapse = ", "))

    for (variable in variables) {
      avg_zscore_col <- paste0("Avg_Zscore_", variable)
      rank_col <- paste0("Rank_", variable)
      df[[rank_col]] <- rank(df[[avg_zscore_col]], na.last = "keep")

      z_lm_col <- paste0("Z_lm_", variable)
      rank_lm_col <- paste0("Rank_lm_", variable)
      df[[rank_lm_col]] <- rank(df[[z_lm_col]], na.last = "keep")
    }
  }

  return(df)
}


main <- function() {
  lapply(names(args$experiments), function(exp_name) {
    exp <- args$experiments[[exp_name]]
    exp_path <- exp$path
    exp_sd <- exp$sd
    out_dir <- file.path(exp_path, "zscores")
    out_dir_qc <- file.path(exp_path, "zscores", "qc")
    dir.create(out_dir, recursive = TRUE, showWarnings = FALSE)
    dir.create(out_dir_qc, recursive = TRUE, showWarnings = FALSE)

    summary_vars <- c("L", "K", "r", "AUC", "delta_bg") # fields to filter and calculate summary stats across
    interaction_vars <- c("L", "K", "r", "AUC") # fields to calculate interaction z-scores
    print_vars <- c("OrfRep", "Plate", "scan", "Col", "Row", "num", "OrfRep", "conc_num", "conc_num_factor",
      "delta_bg_tolerance", "delta_bg", "Gene", "L", "K", "r", "AUC", "NG", "DB")
    
    message("Loading and filtering data for experiment: ", exp_name)
    df <- load_and_process_data(args$easy_results_file, sd = exp_sd) %>%
      update_gene_names(args$sgd_gene_list) %>%
      as_tibble()

    # Filter rows above delta background tolerance
    df_above_tolerance <- df %>% filter(DB == 1)
    df_na <- df %>% mutate(across(all_of(summary_vars), ~ ifelse(DB == 1, NA, .)))
    df_no_zeros <- df_na %>% filter(L > 0)
    
    # Save some constants
    max_conc <- max(df$conc_num)
    l_half_median <- (median(df_above_tolerance$L, na.rm = TRUE)) / 2
    k_half_median <- (median(df_above_tolerance$K, na.rm = TRUE)) / 2

    message("Calculating summary statistics before quality control")
    ss <- calculate_summary_stats(
      df = df,
      variables = summary_vars,
      group_vars = c("conc_num", "conc_num_factor"))
    df_stats <- ss$df_with_stats
    message("Filtering non-finite data")
    df_filtered_stats <- filter_data(df_stats, c("L"), nf = TRUE)

    message("Calculating summary statistics after quality control")
    ss <- calculate_summary_stats(
      df = df_na,
      variables = summary_vars,
      group_vars = c("conc_num", "conc_num_factor"))
    df_na_ss <- ss$summary_stats
    df_na_stats <- ss$df_with_stats
    write.csv(df_na_ss, file = file.path(out_dir, "summary_stats_all_strains.csv"), row.names = FALSE)
    df_na_filtered_stats <- filter_data(df_na_stats, c("L"), nf = TRUE)

    message("Calculating summary statistics after quality control excluding zero values")
    ss <- calculate_summary_stats(
      df = df_no_zeros,
      variables = summary_vars,
      group_vars = c("conc_num", "conc_num_factor"))
    df_no_zeros_stats <- ss$df_with_stats
    df_no_zeros_filtered_stats <- filter_data(df_no_zeros_stats, c("L"), nf = TRUE)

    message("Filtering by 2SD of K")
    df_na_within_2sd_k <- df_na_stats %>%
      filter(K >= (mean_K - 2 * sd_K) & K <= (mean_K + 2 * sd_K))
    df_na_outside_2sd_k <- df_na_stats %>%
      filter(K < (mean_K - 2 * sd_K) | K > (mean_K + 2 * sd_K))

    message("Calculating summary statistics for L within 2SD of K")
    # TODO We're omitting the original z_max calculation, not sure if needed?
    ss <- calculate_summary_stats(df_na_within_2sd_k, "L", group_vars = c("conc_num", "conc_num_factor"))
    # df_na_l_within_2sd_k_stats <- ss$df_with_stats
    write.csv(ss$summary_stats,
      file = file.path(out_dir_qc, "max_observed_L_vals_for_spots_within_2sd_K.csv"),
      row.names = FALSE)
    
    message("Calculating summary statistics for L outside 2SD of K")
    ss <- calculate_summary_stats(df_na_outside_2sd_k, "L", group_vars = c("conc_num", "conc_num_factor"))
    df_na_l_outside_2sd_k_stats <- ss$df_with_stats
    write.csv(ss$summary_stats,
      file = file.path(out_dir, "max_observed_L_vals_for_spots_outside_2sd_K.csv"),
      row.names = FALSE)
    
    # Each plots list corresponds to a file
    l_vs_k_plots <- list(
      list(
        df = df,
        x_var = "L",
        y_var = "K",
        plot_type = "scatter",
        delta_bg_point = TRUE,
        title = "Raw L vs K before quality control",
        color_var = "conc_num_factor",
        error_bar = FALSE,
        legend_position = "right"
      )
    )

    frequency_delta_bg_plots <- list(
      list(
        df = df_filtered_stats,
        x_var = "delta_bg",
        y_var = NULL,
        plot_type = "density",
        title = "Plate analysis by Drug Conc for Delta Background before quality control",
        color_var = "conc_num_factor",
        x_label = "Delta Background",
        y_label = "Density",
        error_bar = FALSE,
        legend_position = "right"),
      list(
        df = df_filtered_stats,
        x_var = "delta_bg",
        y_var = NULL,
        plot_type = "bar",
        title = "Plate analysis by Drug Conc for Delta Background before quality control",
        color_var = "conc_num_factor",
        x_label = "Delta Background",
        y_label = "Count",
        error_bar = FALSE,
        legend_position = "right")
    )

    above_threshold_plots <- list(
      list(
        df = df_above_tolerance,
        x_var = "L",
        y_var = "K",
        plot_type = "scatter",
        delta_bg_point = TRUE,
        title = paste("Raw L vs K for strains above Delta Background threshold of",
          df_above_tolerance$delta_bg_tolerance[[1]], "or above"),
        color_var = "conc_num_factor",
        position = "jitter",
        annotations = list(
          list(
            x = l_half_median,
            y = k_half_median,
            label = paste("# strains above Delta Background tolerance =", nrow(df_above_tolerance))
          )
        ),
        error_bar = FALSE,
        legend_position = "right"
      )
    )

    plate_analysis_plot_configs <- generate_plate_analysis_plot_configs(
      variables = summary_vars,
      df_before = df_filtered_stats,
      df_after = df_na_filtered_stats,
    )

    plate_analysis_boxplot_configs <- generate_plate_analysis_plot_configs(
      variables = summary_vars,
      df_before = df_filtered_stats,
      df_after = df_na_filtered_stats,
      plot_type = "box"
    )

    plate_analysis_no_zeros_plot_configs <- generate_plate_analysis_plot_configs(
      variables = summary_vars,
      stages = c("after"),  # Only after QC
      df_after = df_no_zeros_filtered_stats,
    )

    plate_analysis_no_zeros_boxplot_configs <- generate_plate_analysis_plot_configs(
      variables = summary_vars,
      stages = c("after"),  # Only after QC
      df_after = df_no_zeros_filtered_stats,
      plot_type = "box"
    )

    l_outside_2sd_k_plots <- list(
      list(
        df = df_na_l_outside_2sd_k_stats,
        x_var = "L",
        y_var = "K",
        plot_type = "scatter",
        delta_bg_point = TRUE,
        title = "Raw L vs K for strains falling outside 2SD of the K mean at each Conc",
        color_var = "conc_num_factor",
        position = "jitter",
        legend_position = "right"
      )
    )

    delta_bg_outside_2sd_k_plots <- list(
      list(
        df = df_na_l_outside_2sd_k_stats,
        x_var = "delta_bg",
        y_var = "K",
        plot_type = "scatter",
        gene_point = TRUE,
        title = "Delta Background vs K for strains falling outside 2SD of the K mean at each Conc",
        color_var = "conc_num_factor",
        position = "jitter",
        legend_position = "right"
      )
    )

    message("Generating quality control plots")
    generate_and_save_plots(out_dir_qc, "L_vs_K_before_quality_control", l_vs_k_plots)
    generate_and_save_plots(out_dir_qc, "frequency_delta_background", frequency_delta_bg_plots)
    generate_and_save_plots(out_dir_qc, "L_vs_K_above_threshold", above_threshold_plots)
    generate_and_save_plots(out_dir_qc, "plate_analysis", plate_analysis_plot_configs)
    generate_and_save_plots(out_dir_qc, "plate_analysis_boxplots", plate_analysis_boxplot_configs)
    generate_and_save_plots(out_dir_qc, "plate_analysis_no_zeros", plate_analysis_no_zeros_plot_configs)
    generate_and_save_plots(out_dir_qc, "plate_analysis_no_zeros_boxplots", plate_analysis_no_zeros_boxplot_configs)
    generate_and_save_plots(out_dir_qc, "L_vs_K_for_strains_2SD_outside_mean_K", l_outside_2sd_k_plots)
    generate_and_save_plots(out_dir_qc, "delta_background_vs_K_for_strains_2sd_outside_mean_K", delta_bg_outside_2sd_k_plots)

    # Process background strains
    bg_strains <- c("YDL227C")
    lapply(bg_strains, function(strain) {
      
      message("Processing background strain: ", strain)
      
      # Handle missing data by setting zero values to NA
      # and then removing any rows with NA in L col
      df_bg <- df_na %>%
        filter(OrfRep == strain) %>%
        mutate(
          L = if_else(L == 0, NA, L),
          K = if_else(K == 0, NA, K),
          r = if_else(r == 0, NA, r),
          AUC = if_else(AUC == 0, NA, AUC)
        ) %>%
        filter(!is.na(L))
      
      # Recalculate summary statistics for the background strain
      message("Calculating summary statistics for background strain")
      ss_bg <- calculate_summary_stats(df_bg, summary_vars, group_vars = c("OrfRep", "conc_num", "conc_num_factor"))
      summary_stats_bg <- ss_bg$summary_stats
      # df_bg_stats <- ss_bg$df_with_stats
      write.csv(summary_stats_bg,
        file = file.path(out_dir, paste0("SummaryStats_BackgroundStrains_", strain, ".csv")),
        row.names = FALSE)
      
      # Filter reference and deletion strains
      # Formerly X2_RF (reference strains)
      df_reference <- df_na_stats %>%
        filter(OrfRep == strain) %>%
        mutate(SM = 0)
      
      # Formerly X2 (deletion strains)
      df_deletion <- df_na_stats %>%
        filter(OrfRep != strain) %>%
        mutate(SM = 0)

      # Set the missing values to the highest theoretical value at each drug conc for L
      # Leave other values as 0 for the max/min
      reference_strain <- df_reference %>%
        group_by(conc_num, conc_num_factor) %>%
        mutate(
          max_l_theoretical = max(max_L, na.rm = TRUE),
          L = ifelse(L == 0 & !is.na(L) & conc_num > 0, max_l_theoretical, L),
          SM = ifelse(L >= max_l_theoretical & !is.na(L) & conc_num > 0, 1, SM),
          L = ifelse(L >= max_l_theoretical & !is.na(L) & conc_num > 0, max_l_theoretical, L)) %>%
        ungroup()

      # Ditto for deletion strains
      deletion_strains <- df_deletion %>%
        group_by(conc_num, conc_num_factor) %>%
        mutate(
          max_l_theoretical = max(max_L, na.rm = TRUE),
          L = ifelse(L == 0 & !is.na(L) & conc_num > 0, max_l_theoretical, L),
          SM = ifelse(L >= max_l_theoretical & !is.na(L) & conc_num > 0, 1, SM),
          L = ifelse(L >= max_l_theoretical & !is.na(L) & conc_num > 0, max_l_theoretical, L)) %>%
        ungroup()

      message("Calculating reference strain interaction scores")
      reference_results <- calculate_interaction_scores(reference_strain, max_conc)
      zscores_calculations_reference <- reference_results$calculations
      zscores_interactions_reference <- reference_results$interactions
      zscores_calculations_reference_joined <- reference_results$calculations_joined
      zscores_interactions_reference_joined <- reference_results$interactions_joined

      message("Calculating deletion strain(s) interactions scores")
      deletion_results <- calculate_interaction_scores(deletion_strains, max_conc)
      zscores_calculations <- deletion_results$calculations
      zscores_interactions <- deletion_results$interactions
      zscores_calculations_joined <- deletion_results$calculations_joined
      zscores_interactions_joined <- deletion_results$interactions_joined

      # Writing Z-Scores to file
      write.csv(zscores_calculations_reference, file = file.path(out_dir, "RF_ZScores_Calculations.csv"), row.names = FALSE)
      write.csv(zscores_calculations, file = file.path(out_dir, "ZScores_Calculations.csv"), row.names = FALSE)
      write.csv(zscores_interactions_reference, file = file.path(out_dir, "RF_ZScores_Interaction.csv"), row.names = FALSE)
      write.csv(zscores_interactions, file = file.path(out_dir, "ZScores_Interaction.csv"), row.names = FALSE)

      # Create interaction plots
      message("Generating reference interaction plots")
      reference_plot_configs <- generate_interaction_plot_configs(zscores_interactions_reference_joined, interaction_vars)
      generate_and_save_plots(out_dir, "RF_interactionPlots", reference_plot_configs, grid_layout = list(ncol = 4, nrow = 3))

      message("Generating deletion interaction plots")
      deletion_plot_configs <- generate_interaction_plot_configs(zscores_interactions_joined, interaction_vars)
      generate_and_save_plots(out_dir, "InteractionPlots", deletion_plot_configs, grid_layout = list(ncol = 4, nrow = 3))

      # Define conditions for enhancers and suppressors
      # TODO Add to study config file?
      threshold <- 2
      enhancer_condition_L <- zscores_interactions$Avg_Zscore_L >= threshold
      suppressor_condition_L <- zscores_interactions$Avg_Zscore_L <= -threshold
      enhancer_condition_K <- zscores_interactions$Avg_Zscore_K >= threshold
      suppressor_condition_K <- zscores_interactions$Avg_Zscore_K <= -threshold
      
      # Subset data
      enhancers_L <- zscores_interactions[enhancer_condition_L, ]
      suppressors_L <- zscores_interactions[suppressor_condition_L, ]
      enhancers_K <- zscores_interactions[enhancer_condition_K, ]
      suppressors_K <- zscores_interactions[suppressor_condition_K, ]
      
      # Save enhancers and suppressors
      message("Writing enhancer/suppressor csv files")
      write.csv(enhancers_L, file = file.path(out_dir, "ZScores_Interaction_Deletion_Enhancers_L.csv"), row.names = FALSE)
      write.csv(suppressors_L, file = file.path(out_dir, "ZScores_Interaction_Deletion_Suppressors_L.csv"), row.names = FALSE)
      write.csv(enhancers_K, file = file.path(out_dir, "ZScores_Interaction_Deletion_Enhancers_K.csv"), row.names = FALSE)
      write.csv(suppressors_K, file = file.path(out_dir, "ZScores_Interaction_Deletion_Suppressors_K.csv"), row.names = FALSE)
      
      # Combine conditions for enhancers and suppressors
      enhancers_and_suppressors_L <- zscores_interactions[enhancer_condition_L | suppressor_condition_L, ]
      enhancers_and_suppressors_K <- zscores_interactions[enhancer_condition_K | suppressor_condition_K, ]
      
      # Save combined enhancers and suppressors
      write.csv(enhancers_and_suppressors_L,
        file = file.path(out_dir, "ZScores_Interaction_Deletion_Enhancers_and_Suppressors_L.csv"), row.names = FALSE)
      write.csv(enhancers_and_suppressors_K,
        file = file.path(out_dir, "ZScores_Interaction_Deletion_Enhancers_and_Suppressors_K.csv"), row.names = FALSE)
      
      # Handle linear model based enhancers and suppressors
      lm_threshold <- 2
      enhancers_lm_L <- zscores_interactions[zscores_interactions$Z_lm_L >= lm_threshold, ]
      suppressors_lm_L <- zscores_interactions[zscores_interactions$Z_lm_L <= -lm_threshold, ]
      enhancers_lm_K <- zscores_interactions[zscores_interactions$Z_lm_K >= lm_threshold, ]
      suppressors_lm_K <- zscores_interactions[zscores_interactions$Z_lm_K <= -lm_threshold, ]
      
      # Save linear model based enhancers and suppressors
      message("Writing linear model enhancer/suppressor csv files")
      write.csv(enhancers_lm_L,
        file = file.path(out_dir, "ZScores_Interaction_Deletion_Enhancers_L_lm.csv"), row.names = FALSE)
      write.csv(suppressors_lm_L,
        file = file.path(out_dir, "ZScores_Interaction_Deletion_Suppressors_L_lm.csv"), row.names = FALSE)
      write.csv(enhancers_lm_K,
        file = file.path(out_dir, "ZScores_Interaction_Deletion_Enhancers_K_lm.csv"), row.names = FALSE)
      write.csv(suppressors_lm_K,
        file = file.path(out_dir, "ZScores_Interaction_Deletion_Suppressors_K_lm.csv"), row.names = FALSE)

      message("Generating rank plots")
      zscores_interactions_joined_filtered <- filter_data(
        df = zscores_interactions_joined,
        variables = interaction_vars,
        missing = TRUE,
        adjust = TRUE,
        rank = TRUE)
      rank_plot_configs <- generate_rank_plot_configs(
        df = zscores_interactions_joined_filtered,
        variables = interaction_vars,
        is_lm = FALSE
      )
      generate_and_save_plots(out_dir = out_dir, file_name = "RankPlots",
        plot_configs = rank_plot_configs, grid_layout = list(ncol = 3, nrow = 2))

      message("Generating ranked linear model plots")
      rank_lm_plot_configs <- generate_rank_plot_configs(
        df = zscores_interactions_joined_filtered,
        variables = interaction_vars,
        is_lm = TRUE
      )
      generate_and_save_plots(out_dir = out_dir, file_name = "RankPlots_lm",
        plot_configs = rank_lm_plot_configs, grid_layout = list(ncol = 3, nrow = 2))

      message("Filtering and reranking plots")
      # Filter out rows where both Z_lm_L and Avg_Zscore_L are NA
      zscores_interactions_filtered <- zscores_interactions_joined %>%
        filter(!is.na(Z_lm_L) | !is.na(Avg_Zscore_L))

      # Formerly X_NArm
      zscores_interactions_filtered <- zscores_interactions_filtered %>%
        mutate(
          lm_R_squared_L = if (n() > 1) summary(lm(Z_lm_L ~ Avg_Zscore_L))$r.squared else NA,
          lm_R_squared_K = if (n() > 1) summary(lm(Z_lm_K ~ Avg_Zscore_K))$r.squared else NA,
          lm_R_squared_r = if (n() > 1) summary(lm(Z_lm_r ~ Avg_Zscore_r))$r.squared else NA,
          lm_R_squared_AUC = if (n() > 1) summary(lm(Z_lm_AUC ~ Avg_Zscore_AUC))$r.squared else NA,
          Overlap = case_when(
            Z_lm_L >= 2 & Avg_Zscore_L >= 2 ~ "Deletion Enhancer Both",
            Z_lm_L <= -2 & Avg_Zscore_L <= -2 ~ "Deletion Suppressor Both",
            Z_lm_L >= 2 & Avg_Zscore_L <= 2 ~ "Deletion Enhancer lm only",
            Z_lm_L <= 2 & Avg_Zscore_L >= 2 ~ "Deletion Enhancer Avg Zscore only",
            Z_lm_L <= -2 & Avg_Zscore_L >= -2 ~ "Deletion Suppressor lm only",
            Z_lm_L >= -2 & Avg_Zscore_L <= -2 ~ "Deletion Suppressor Avg Zscore only",
            Z_lm_L >= 2 & Avg_Zscore_L <= -2 ~ "Deletion Enhancer lm, Deletion Suppressor Avg Z score",
            Z_lm_L <= -2 & Avg_Zscore_L >= 2 ~ "Deletion Suppressor lm, Deletion Enhancer Avg Z score",
            TRUE ~ "No Effect"
          )
        )

      # Re-rank
      zscores_interactions_filtered <- filter_data(
        df = zscores_interactions_filtered,
        variables = interaction_vars,
        missing = TRUE, # TODO what I'm currently having issues with
        rank = TRUE
      )

      rank_plot_filtered_configs <- generate_rank_plot_configs(
        df = zscores_interactions_filtered,
        variables = interaction_vars,
        is_lm = FALSE
      )

      message("Generating filtered ranked plots")
      generate_and_save_plots(
        out_dir = out_dir,
        file_name = "RankPlots_na_rm",
        plot_configs = rank_plot_filtered_configs,
        grid_layout = list(ncol = 3, nrow = 2))

      message("Generating filtered ranked linear model plots")
      rank_plot_lm_filtered_configs <- generate_rank_plot_configs(
        df = zscores_interactions_filtered,
        variables = interaction_vars,
        is_lm = TRUE
      )
      generate_and_save_plots(
        out_dir = out_dir,
        file_name = "RankPlots_lm_na_rm",
        plot_configs = rank_plot_lm_filtered_configs,
        grid_layout = list(ncol = 3, nrow = 2))

      message("Generating correlation plots")
      correlation_plot_configs <- generate_correlation_plot_configs(zscores_interactions_filtered)
      generate_and_save_plots(
        out_dir = out_dir,
        file_name = "Avg_Zscore_vs_lm_NA_rm",
        plot_configs = correlation_plot_configs,
        grid_layout = list(ncol = 2, nrow = 2))
    })
  })
}
main()