-
Notifications
You must be signed in to change notification settings - Fork 5
Commit
This commit does not belong to any branch on this repository, and may belong to a fork outside of the repository.
- Loading branch information
Reed Nelson
committed
Oct 6, 2024
1 parent
7c27bf0
commit 0712960
Showing
6 changed files
with
157 additions
and
114 deletions.
There are no files selected for viewing
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
Loading
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.
File renamed without changes
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| Original file line number | Diff line number | Diff line change |
|---|---|---|
| @@ -1,106 +1,145 @@ | ||
| plot_alignment <- function(g_filepath, h_filepath, alignment_filepath, output_filepath, th_align, zero_degree, vertex_label_value, size_aligned) { | ||
|
|
||
| # Load adjacency matrices and alignment matrix | ||
| adj_G <- as.matrix(read.csv(g_filepath, header = TRUE, row.names = 1, check.names = FALSE)) | ||
| adj_H <- as.matrix(read.csv(h_filepath, header = TRUE, row.names = 1, check.names = FALSE)) | ||
| alignment_GH <- as.matrix(read.csv(alignment_filepath, header = TRUE, row.names = 1, check.names = FALSE)) | ||
|
|
||
| g_size <- nrow(adj_G) | ||
| h_size <- nrow(adj_H) | ||
| adj_G <- adj_G[1:g_size, 1:g_size] | ||
| adj_H <- adj_H[1:h_size, 1:h_size] | ||
| alignment_GH <- alignment_GH[1:g_size, 1:h_size] | ||
|
|
||
| # Create graph objects with unique vertex names | ||
| graph_G <- graph_from_adjacency_matrix(adj_G, mode = "undirected") | ||
| V(graph_G)$name <- rownames(adj_G) | ||
|
|
||
| graph_H <- graph_from_adjacency_matrix(adj_H, mode = "undirected") | ||
| V(graph_H)$name <- rownames(adj_H) | ||
|
|
||
| # Identify aligned nodes | ||
| alignment_indices <- which(alignment_GH >= th_align, arr.ind = TRUE) | ||
| aligned_nodes_G <- V(graph_G)$name[alignment_indices[, 1]] | ||
| aligned_nodes_H <- V(graph_H)$name[alignment_indices[, 2]] | ||
|
|
||
| # Optionally remove nodes with degree 0, except aligned nodes | ||
| zero_deg_nodes_G <- V(graph_G)[degree(graph_G) < zero_degree & !(V(graph_G)$name %in% aligned_nodes_G)] | ||
| zero_deg_nodes_H <- V(graph_H)[degree(graph_H) < zero_degree & !(V(graph_H)$name %in% aligned_nodes_H)] | ||
| graph_G <- delete.vertices(graph_G, zero_deg_nodes_G) | ||
| graph_H <- delete.vertices(graph_H, zero_deg_nodes_H) | ||
|
|
||
| # Re-calculate the aligned nodes after potential removal | ||
| aligned_indices_G <- match(aligned_nodes_G, V(graph_G)$name) | ||
| aligned_indices_H <- match(aligned_nodes_H, V(graph_H)$name) | ||
|
|
||
| # Filter out NA values from aligned indices | ||
| valid_indices <- which(!is.na(aligned_indices_G) & !is.na(aligned_indices_H)) | ||
| aligned_indices_G <- aligned_indices_G[valid_indices] | ||
| aligned_indices_H <- aligned_indices_H[valid_indices] | ||
|
|
||
| # Create a combined graph for alignment visualization | ||
| combined_graph <- graph.disjoint.union(graph_G, graph_H) | ||
|
|
||
| # Define layout for combined graph with circular layout | ||
| layout_G <- layout_in_circle(graph_G) | ||
| layout_H <- layout_in_circle(graph_H) | ||
|
|
||
| png(output_filepath, width = 1600, height = 1200, res = 120) | ||
|
|
||
| # Adjust the layout to place aligned nodes directly opposite each other | ||
| adjust_opposite_positions <- function(layout_G, layout_H, indices_G, indices_H) { | ||
| num_nodes <- length(indices_G) | ||
| angle_step <- pi / (num_nodes + 1) | ||
| start_angle <- pi / 2 | ||
|
|
||
| for (i in seq_along(indices_G)) { | ||
| angle <- start_angle - (i * angle_step) | ||
| layout_G[indices_G[i], ] <- c(cos(angle), sin(angle)) | ||
| layout_H[indices_H[i], ] <- c(-cos(angle), sin(angle)) | ||
| plot_alignment <- function( | ||
| g_filepath, | ||
| h_filepath, | ||
| alignment_filepath, | ||
| output_filepath, | ||
| st = 0, | ||
| hide_singletons = FALSE | ||
| ) { | ||
| # # Load necessary libraries | ||
| # required_packages <- c("igraph", "ggraph", "ggplot2", "readr", "dplyr") | ||
| # installed_packages <- rownames(installed.packages()) | ||
| # for (pkg in required_packages) { | ||
| # if (!pkg %in% installed_packages) { | ||
| # install.packages(pkg, dependencies = TRUE) | ||
| # } | ||
| # } | ||
|
|
||
| library(igraph) | ||
| library(ggraph) | ||
| library(ggplot2) | ||
| library(readr) | ||
| library(dplyr) | ||
|
|
||
| # Helper function to read adjacency matrix | ||
| read_adjacency <- function(filepath) { | ||
| df <- read_csv(filepath, show_col_types = FALSE) | ||
| mat <- as.matrix(df[, -1]) | ||
| rownames(mat) <- df[[1]] | ||
| return(mat) | ||
| } | ||
| return(list(layout_G, layout_H)) | ||
| } | ||
|
|
||
| layouts <- adjust_opposite_positions(layout_G, layout_H, aligned_indices_G, aligned_indices_H) | ||
| layout_G <- layouts[[1]] | ||
| layout_H <- layouts[[2]] | ||
|
|
||
| # Reduce the separation distance between G and H networks | ||
| layout_G <- layout_G - cbind(rep(2, nrow(layout_G)), rep(0, nrow(layout_G))) | ||
| layout_H <- layout_H + cbind(rep(2, nrow(layout_H)), rep(0, nrow(layout_H))) | ||
|
|
||
| # Combine layouts | ||
| layout_combined <- rbind(layout_G, layout_H) | ||
|
|
||
| # Add edges based on alignment matrix GH | ||
| for (i in seq_along(aligned_indices_G)) { | ||
| combined_graph <- add_edges(combined_graph, c(aligned_indices_G[i], aligned_indices_H[i] + vcount(graph_G))) | ||
| } | ||
|
|
||
| # Define colors for vertices and edges | ||
| vertex_colors <- c(rep("lightblue", vcount(graph_G)), rep("lightgreen", vcount(graph_H))) | ||
| vertex_colors[aligned_indices_G] <- "dodgerblue" | ||
| vertex_colors[aligned_indices_H + vcount(graph_G)] <- "green" | ||
| edge_colors <- c(rep("black", ecount(graph_G) + ecount(graph_H)), rep("purple", length(aligned_indices_G))) | ||
|
|
||
| # Define sizes and label sizes for vertices | ||
| vertex_sizes <- rep(1, vcount(combined_graph)) | ||
| vertex_sizes[c(aligned_indices_G, aligned_indices_H + vcount(graph_G))] <- size_aligned | ||
| vertex_label_sizes <- rep(vertex_label_value, vcount(combined_graph)) | ||
| vertex_label_sizes[c(aligned_indices_G, aligned_indices_H + vcount(graph_G))] <- 0.8 | ||
|
|
||
| # Plot the combined graph with alignment edges | ||
| plot(combined_graph, | ||
| layout = layout_combined, | ||
| vertex.label = V(combined_graph)$name, # Add vertex labels | ||
| vertex.size = vertex_sizes, # Adjust vertex size for better visualization | ||
| vertex.label.cex = vertex_label_sizes, # Adjust label size for better readability | ||
| vertex.color = vertex_colors, | ||
| edge.color = edge_colors, | ||
| edge.width = c(rep(1, ecount(graph_G) + ecount(graph_H)), rep(2, length(aligned_indices_G))), | ||
| main = "Result of MiNAA", | ||
| asp = 0 # Set aspect ratio to ensure circular layout | ||
| ) | ||
|
|
||
| dev.off() | ||
| } | ||
|
|
||
| # Read adjacency matrices | ||
| g_mat <- read_adjacency(g_filepath) | ||
| h_mat <- read_adjacency(h_filepath) | ||
|
|
||
| # Create igraph objects (assuming undirected graphs; adjust 'mode' if needed) | ||
| g_graph <- graph_from_adjacency_matrix(g_mat, mode = "undirected", diag = FALSE) | ||
| h_graph <- graph_from_adjacency_matrix(h_mat, mode = "undirected", diag = FALSE) | ||
|
|
||
| # Read alignment matrix | ||
| alignment_df <- read_csv(alignment_filepath, show_col_types = FALSE) | ||
| alignment_mat <- as.matrix(alignment_df[, -1]) | ||
| rownames(alignment_mat) <- alignment_df[[1]] | ||
|
|
||
| # Extract alignment pairs with similarity > the similarity threshold | ||
| alignment_pairs <- which(alignment_mat > st, arr.ind = TRUE) | ||
| align_df <- data.frame( | ||
| g_node = rownames(alignment_mat)[alignment_pairs[, 1]], | ||
| h_node = colnames(alignment_mat)[alignment_pairs[, 2]], | ||
| similarity = alignment_mat[alignment_pairs] | ||
| ) | ||
|
|
||
| if (hide_singletons) { | ||
| # Remove nodes in G and H with degree 0 | ||
| deg_g <- degree(g_graph) | ||
| non_single_g <- names(deg_g[deg_g > 0]) | ||
| g_graph <- induced_subgraph(g_graph, vids = non_single_g) | ||
|
|
||
| deg_h <- degree(h_graph) | ||
| non_single_h <- names(deg_h[deg_h > 0]) | ||
| h_graph <- induced_subgraph(h_graph, vids = non_single_h) | ||
|
|
||
| # Update alignment_df to include only existing nodes after subsetting | ||
| align_df <- align_df %>% | ||
| filter(g_node %in% V(g_graph)$name & h_node %in% V(h_graph)$name) | ||
| } | ||
|
|
||
| # Rename nodes to ensure uniqueness | ||
| V(g_graph)$name <- paste0("G_", V(g_graph)$name) | ||
| V(h_graph)$name <- paste0("H_", V(h_graph)$name) | ||
|
|
||
| combined_graph <- disjoint_union(g_graph, h_graph) | ||
|
|
||
| # Prepare alignment edges | ||
| if (nrow(align_df) > 0) { | ||
| align_edges <- align_df %>% | ||
| mutate( | ||
| from = paste0("G_", g_node), | ||
| to = paste0("H_", h_node) | ||
| ) | ||
| } else { | ||
| align_edges <- data.frame(from = character(0), to = character(0), similarity = numeric(0)) | ||
| } | ||
|
|
||
| # Create a layout with G on the left and H on the right | ||
| set.seed(123) # For reproducibility | ||
| layout_g <- layout_with_fr(g_graph) | ||
| layout_h <- layout_with_fr(h_graph) | ||
|
|
||
| # Shift H layout to the right of G | ||
| shift_x <- max(layout_g[, 1]) - min(layout_h[, 1]) + 5 | ||
| layout_h[, 1] <- layout_h[, 1] + shift_x | ||
|
|
||
| # Combine layouts | ||
| combined_layout <- rbind(layout_g, layout_h) | ||
| V(combined_graph)$x <- combined_layout[, 1] | ||
| V(combined_graph)$y <- combined_layout[, 2] | ||
|
|
||
| # Prepare data for alignment edges | ||
| if (nrow(align_edges) > 0) { | ||
| align_coords <- align_edges %>% | ||
| mutate( | ||
| from_x = V(combined_graph)$x[match(from, V(combined_graph)$name)], | ||
| from_y = V(combined_graph)$y[match(from, V(combined_graph)$name)], | ||
| to_x = V(combined_graph)$x[match(to, V(combined_graph)$name)], | ||
| to_y = V(combined_graph)$y[match(to, V(combined_graph)$name)] | ||
| ) | ||
| } | ||
|
|
||
| # Plot # | ||
|
|
||
| # Adjust the plot size based on the number of nodes | ||
| num_nodes <- vcount(combined_graph) | ||
| plot_width <- max(2000, 100 + 20 * num_nodes) | ||
| plot_height <- max(1000, 100 + 10 * num_nodes) | ||
|
|
||
| png(filename = output_filepath, width = plot_width, height = plot_height, res = 150) | ||
|
|
||
| p <- ggraph(combined_graph, layout = "manual", x = V(combined_graph)$x, y = V(combined_graph)$y) + | ||
| # Internal edges within G and H | ||
| geom_edge_link(aes(alpha = 0.5), color = "grey", linewidth = 0.5) + | ||
| # G nodes in red, H in blue | ||
| geom_node_point(aes(color = ifelse(startsWith(name, "G_"), "G", "H")), size = 3) + | ||
| # Node labels without the prefix | ||
| geom_node_text(aes(label = gsub("^[GH]_", "", name)), repel = TRUE, size = 3) | ||
| # Color mapping | ||
| scale_color_manual(values = c("G" = "red", "H" = "blue")) + | ||
| theme_void() + | ||
| theme(legend.position = "none") | ||
|
|
||
| if (nrow(align_edges) > 0) { | ||
| # Add alignment edges | ||
| p <- p + | ||
| geom_segment( | ||
| data = align_coords, | ||
| aes(x = from_x, y = from_y, xend = to_x, yend = to_y), | ||
| color = "purple", linewidth = 0.5, alpha = 0.6 | ||
| ) | ||
| } | ||
|
|
||
| print(p) | ||
| dev.off() | ||
|
|
||
| message("Plot saved to ", output_filepath) | ||
| } |