bryan/hartman-server
1
0
Fork 0
Code Issues Pull Requests Releases Wiki Activity

Reformat rank plot configs

Browse Source
This commit is contained in:
Bryan Roessler
2024-09-17 15:06:27 -04:00
parent 012c5e1d69
commit 2ddba41950
1 changed files with 133 additions and 141 deletions
Download Patch File Download Diff File Expand all files Collapse all files

256
qhtcp-workflow/apps/r/calculate_interaction_zscores.R
View File

@@ -721,140 +721,132 @@ generate_interaction_plot_configs <- function(df, variables) {
} }
generate_rank_plot_configs <- function(df_filtered, variables, is_lm = FALSE) { generate_rank_plot_configs <- function(df_filtered, variables, is_lm = FALSE) {
# Define SD bands
sd_bands <- c(1, 2, 3)
# Initialize list to store plot configurations sd_bands <- c(1, 2, 3)
configs <- list()
# SD-based plots for L and K configs <- list()
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(variable, "_Rank_lm")
zscore_var <- paste0("Z_lm_", variable)
y_label <- paste("Int Z score", variable)
} else {
rank_var <- paste0(variable, "_Rank")
zscore_var <- paste0("Avg_Zscore_", variable)
y_label <- paste("Avg Z score", variable)
}
# Calculate counts for annotations # SD-based plots for L and K
num_enhancers <- sum(df_filtered[[zscore_var]] >= sd_band, na.rm = TRUE) for (variable in c("L", "K")) {
num_suppressors <- sum(df_filtered[[zscore_var]] <= -sd_band, na.rm = TRUE) for (sd_band in sd_bands) {
# Determine columns based on whether it's lm or not
if (is_lm) {
rank_var <- paste0(variable, "_Rank_lm")
zscore_var <- paste0("Z_lm_", variable)
y_label <- paste("Int Z score", variable)
} else {
rank_var <- paste0(variable, "_Rank")
zscore_var <- paste0("Avg_Zscore_", variable)
y_label <- paste("Avg Z score", variable)
}
# Annotated Plot Configuration num_enhancers <- sum(df_filtered[[zscore_var]] >= sd_band, na.rm = TRUE)
configs[[length(configs) + 1]] <- list( num_suppressors <- sum(df_filtered[[zscore_var]] <= -sd_band, na.rm = TRUE)
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 # Annotated plot configuration
configs[[length(configs) + 1]] <- list( configs[[length(configs) + 1]] <- list(
df = df_filtered, df = df_filtered,
x_var = rank_var, x_var = rank_var,
y_var = zscore_var, y_var = zscore_var,
plot_type = "scatter", plot_type = "scatter",
title = paste(y_label, "vs. Rank for", variable, "above", sd_band, "SD No Annotations"), title = paste(y_label, "vs. Rank for", variable, "above", sd_band, "SD"),
sd_band = sd_band, sd_band = sd_band,
annotations = NULL, annotations = list(
sd_band_values = sd_band, list(
shape = 3, x = median(df_filtered[[rank_var]], na.rm = TRUE),
size = 0.1, y = 10,
y_label = y_label, label = paste("Deletion Enhancers =", num_enhancers)
x_label = "Rank", ),
legend_position = "none" 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 # Avg ZScore and Rank Avg ZScore Plots for r, L, K, and AUC
for (variable in variables) { for (variable in variables) {
for (plot_type in c("Avg_Zscore_vs_lm", "Rank_Avg_Zscore_vs_lm")) { for (plot_type in c("Avg_Zscore_vs_lm", "Rank_Avg_Zscore_vs_lm")) {
# Define x and y variables based on plot type # Define specific variables based on plot type
if (plot_type == "Avg_Zscore_vs_lm") { if (plot_type == "Avg_Zscore_vs_lm") {
x_var <- paste0("Avg_Zscore_", variable) x_var <- paste0("Avg_Zscore_", variable)
y_var <- paste0("Z_lm_", variable) y_var <- paste0("Z_lm_", variable)
title_suffix <- paste("Avg Zscore vs 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(variable, "_Rank")
y_var <- paste0(variable, "_Rank_lm")
title_suffix <- paste("Rank Avg Zscore vs lm", variable)
rectangles <- NULL
}
# Add rectangles configuration for Avg_Zscore_vs_lm # Fit linear model
rectangles <- list( lm_fit <- lm(df_filtered[[y_var]] ~ df_filtered[[x_var]], data = df_filtered)
list(
xmin = -2, xmax = 2, ymin = -2, ymax = 2,
fill = NA, color = "grey20", alpha = 0.1
)
)
} else {
x_var <- paste0(variable, "_Rank")
y_var <- paste0(variable, "_Rank_lm")
title_suffix <- paste("Rank Avg Zscore vs lm", variable)
rectangles <- NULL
}
# Fit linear model # Check for perfect fit
lm_fit <- lm(df_filtered[[y_var]] ~ df_filtered[[x_var]], data = df_filtered) if (summary(lm_fit)$sigma == 0) {
next # Skip this iteration if the fit is perfect
}
# Check for perfect fit r_squared <- summary(lm_fit)$r.squared
if (summary(lm_fit)$sigma == 0) {
next # Skip this iteration if the fit is perfect
}
# Calculate R-squared configs[[length(configs) + 1]] <- list(
r_squared <- summary(lm_fit)$r.squared df = df_filtered,
x_var = x_var,
# Plot Configuration y_var = y_var,
configs[[length(configs) + 1]] <- list( plot_type = "scatter",
df = df_filtered, title = title_suffix,
x_var = x_var, annotations = list(
y_var = y_var, list(
plot_type = "scatter", x = 0,
title = title_suffix, y = 0,
annotations = list( label = paste("R-squared =", round(r_squared, 2))
list( )
x = 0, ),
y = 0, sd_band_values = NULL, # Not applicable
label = paste("R-squared =", round(r_squared, 2)) shape = 3,
) size = 0.1,
), add_smooth = TRUE,
sd_band_values = NULL, # Not applicable lm_line = list(intercept = coef(lm_fit)[1], slope = coef(lm_fit)[2]),
shape = 3, legend_position = "right",
size = 0.1, color_var = "Overlap",
add_smooth = TRUE, x_label = x_var,
lm_line = list(intercept = coef(lm_fit)[1], slope = coef(lm_fit)[2]), y_label = y_var,
legend_position = "right", rectangles = rectangles # Add rectangles configuration
color_var = "Overlap", )
x_label = x_var,
y_label = y_var,
rectangles = rectangles # Add rectangles configuration
)
}
} }
}
return(configs) return(configs)
} }
generate_correlation_plot_configs <- function(df) { generate_correlation_plot_configs <- function(df) {
@@ -1247,16 +1239,16 @@ main <- function() {
) )
) )
message("Generating quality control plots") # 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, "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, "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, "L_vs_K_above_threshold", above_threshold_plots)
generate_and_save_plots(out_dir_qc, "plate_analysis", plate_analysis_plots) # generate_and_save_plots(out_dir_qc, "plate_analysis", plate_analysis_plots)
generate_and_save_plots(out_dir_qc, "plate_analysis_boxplots", plate_analysis_boxplots) # generate_and_save_plots(out_dir_qc, "plate_analysis_boxplots", plate_analysis_boxplots)
generate_and_save_plots(out_dir_qc, "plate_analysis_no_zeros", plate_analysis_no_zeros_plots) # generate_and_save_plots(out_dir_qc, "plate_analysis_no_zeros", plate_analysis_no_zeros_plots)
generate_and_save_plots(out_dir_qc, "plate_analysis_no_zeros_boxplots", plate_analysis_no_zeros_boxplots) # generate_and_save_plots(out_dir_qc, "plate_analysis_no_zeros_boxplots", plate_analysis_no_zeros_boxplots)
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, "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) # 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 # Process background strains
bg_strains <- c("YDL227C") bg_strains <- c("YDL227C")