R/02_plot_logistic_SVD_kmeans.R

In blind-contours/ChemLogic: Pipeline of methods that take in chemical fingerprint and molecular physical information and creates interpretable rules based on this information which leads to nuclear receptor binding

#' Plot the first two PCs from logistic SVD for molecular features and color by k-means results
#'
#' This function takes in a binary matrix, assuming that the variable names include the fingerprint name such as "Pubchem"
#'
#' @param log_SVD_results Results from the SVD to get data from for plotting
#' @param receptor_type type of receptor for labeling plots
#' @param k number of groups from kmeans from after viewing elbow plot or other methods of determining groups
#' @return Plot of SVD results colored by kmeans groups
#' @export
#'
plot_SVD_kmeans <- function(receptor_type,
                            log_SVD_results,
                            binary_data,
                            k = 4,
                            outcome_label,
                            endo_ligand = FALSE,
                            endo_ligand_id,
                            additional_label = TRUE,
                            binding_label = "ASSAY_OUTCOME") {
  #browser()
  cluster_data_actives <- binary_data %>% dplyr::filter(!!outcome_label == 1)

  kmeans_groups <- kmeans(log_SVD_results$A,
                          k,
                          nstart = 50,
                          iter.max = 15)

  groups <- kmeans_groups$cluster

  if (endo_ligand) {
    groups_endo_idx <- groups
    idx <- match(endo_ligand_id, cluster_data_actives$Name)
    groups_endo_idx[idx] <- 20
  }

  plot_w_kmeans <- plot(log_SVD_results,
                        type = "scores") +
    geom_point(aes(colour = as.factor(groups_endo_idx))) +
    ggtitle(paste("Positive Binding for ", receptor_type, sep = ""))

  data_negatives <- binary_data %>%
    dplyr::filter(!!outcome_label == 0)

  cluster_data_actives$log_SVM_group <- groups
  cluster_data_actives$PC_1 <- log_SVD_results$A[, 1]
  cluster_data_actives$PC_2 <- log_SVD_results$A[, 2]

  data_negatives$log_SVM_group <- 0
  data_negatives$log_SVM_group <- 0

  data_negatives$PC_1 <- NA
  data_negatives$PC_2 <- NA

  data_with_PCs_kmeans_groups <- rbind(cluster_data_actives,
                                       data_negatives)

  if (additional_label) {
    plot_outcome <- plot(log_SVD_results, type = "scores") +
      geom_point(aes(colour = as.factor(cluster_data_actives[,binding_label]))) +
      ggtitle(paste("Positive Binding for ", receptor_type, sep = ""))
  } else {
    plot_outcome <- NA
  }

  if (endo_ligand) {
    return(list(plot_kmeans = plot_w_kmeans,
                plot_outcome_label = plot_outcome,
                updated_data = data_with_PCs_kmeans_groups,
                endo_ligand_grp_included = groups_endo_idx))
  } else {
    list(plot_kmeans = plot_w_kmeans,
         plot_outcome_label = plot_outcome,
         updated_data = data_with_PCs_kmeans_groups)
  }
}