version2.4

DESCRIPTION

@@ -1,7 +1,7 @@
 Package: SPARKLE
 Type: Package
 Title: Single-cell Phenotype Association Research Kit for Large-scale Evaluation
-Version: 0.2.3
+Version: 0.2.4
 Authors@R: c(
  person("Chen", "Xi", ,email = "[email protected]", role = c("aut", "cre")),
  person("Ding", "Yang", ,email = "[email protected]", role = "ctb"),

R/f13.1.Moderation_Analysis_kernal.R

@@ -82,7 +82,7 @@ moderation_analysis_auto <- function(cwas.data,Best_model, X.cell=NULL,X.gene=NU
  names(Best_model[["Chosen_model"]]) <- gsub("-", "_", names(Best_model[["Chosen_model"]]) )
  X.cell.fomula <- as.character(Best_model[["Chosen_model"]][[X.cell]]@call[["formula"]])[3]
  X.cell.fomula <- strsplit(X.cell.fomula, " \\+ ")[[1]]
- #X.cell.fomula <-replace_double_pipe(X.cell.fomula )
+ X.cell.fomula <-replace_double_pipe(X.cell.fomula )
  X.cell.fomula.result <- formula_process_string_vector(X.cell.fomula,ratename = X.cell,clean = Cleaninfo)
 
  return(X.cell.fomula.result)

R/f7.2.cwas_allmodel_heatmap.R

@@ -1,28 +1,28 @@
 #' Generate a heatmap of AIC values for multiple models
-#' 
+#'
 #' This function takes a list of model dataframes and generates a heatmap
 #' to visualize the AIC values across different models.
-#' 
-#' @param mdlist A list of model dataframes. Each element in the list corresponds 
+#'
+#' @param mdlist A list of model dataframes. Each element in the list corresponds
 #' to a different cell type or model type.
-#' @param method Character string specifying the method to use for ranking. 
+#' @param method Character string specifying the method to use for ranking.
 #' Currently supports "rank" (default) and "AIC".
-#' 
+#'
 #' @details
 #' The function processes each model dataframe in the input list (\code{mdlist}),
 #' extracts relevant columns (formula, AIC, Pvalue, rank), and combines them into
 #' a single dataframe (\code{alldata}). It then cleans unnecessary columns,
-#' orders the data, and prepares it for heatmap plotting. P-values are used to 
+#' orders the data, and prepares it for heatmap plotting. P-values are used to
 #' annotate the heatmap cells based on significance levels.
-#' 
+#'
 #' @examples
 #' # Example usage:
 #' # cwas_allmodel_heatmap(list(model_data1, model_data2))
-#' 
+#'
 #' @import pheatmap
 #' @export
-#' 
- 
+#'
+
 cwas_allmodel_heatmap <- function (mdlist,method=c("rank")) {
  i = 1
  celltype <- names(mdlist)
@@ -39,61 +39,61 @@ cwas_allmodel_heatmap <- function (mdlist,method=c("rank")) {
  colnames(mydata) <- paste0(celltype[i], ".", colnames(mydata))
  mydata$name <- mydata[, 1]
  alldata <- merge(alldata, mydata, by = "name", all = T)
- print(i)
+ #print(i)
  }
  alldata_cleaned <- alldata[, !grepl("celltype", names(alldata))]
  alldata_cleaned <- alldata_cleaned[, !grepl("model", names(alldata_cleaned))]
  alldata_cleaned <- alldata_cleaned[, !grepl("formula", names(alldata_cleaned))]
  alldata_cleaned <- alldata_cleaned[, !grepl("id", names(alldata_cleaned))]
  alldata_cleaned$formula <- alldata[, 2]
  alldata_cleaned$modelID <- alldata[, 3]
- alldata_cleaned <- alldata_cleaned[order(alldata_cleaned$modelID), 
+ alldata_cleaned <- alldata_cleaned[order(alldata_cleaned$modelID),
  ]
  alldata <- alldata[order(alldata[, 3]), ]
  allformula <- alldata$name
- pvalue_columns <- grep("Pvalue", names(alldata_cleaned), 
+ pvalue_columns <- grep("Pvalue", names(alldata_cleaned),
  value = TRUE)
  AICrank_columns <- grep(method, names(alldata_cleaned), value = TRUE)
  pvalue_data <- alldata_cleaned[c("name", pvalue_columns)]
  heatmap_data <- alldata_cleaned[c("name", AICrank_columns)]
- 
+
  ## 绘制热图
- 
+
  heatmap_data2 <- heatmap_data
  heatmap_data2[, -1] <- lapply(heatmap_data2[, -1], as.numeric)
  max_value <- max(heatmap_data2[, -1], na.rm = TRUE)
- 
+
  # 将 heatmap_data2 中的所有 NA 值替换成最大值
  heatmap_data2[is.na(heatmap_data2)] <- max_value
  heatmap_data2$name <- gsub("binomial link function: ", "", heatmap_data2$name)
  row.names(heatmap_data2) <- heatmap_data2$name
  colnames(heatmap_data2) <- gsub(".rank", "", colnames(heatmap_data2))
  colnames(heatmap_data2) <- gsub(".AIC", "", colnames(heatmap_data2))
  heatmap_data2$name=NULL
- 
+
  pvalue_matrix <- pvalue_data
  pvalue_matrix$name=NULL
  rownames(pvalue_matrix) <- rownames(heatmap_data2)
  colnames(pvalue_matrix) <- colnames(heatmap_data2)
- 
+
  # 定义p值注释
  annotation_matrix <- ifelse(pvalue_matrix < 0.001, "***",
  ifelse(pvalue_matrix < 0.01, "**",
  ifelse(pvalue_matrix < 0.05, "*", "")))
  annotation_matrix[is.na(annotation_matrix)] <- "NA"
- 
- 
- pheatmap::pheatmap(heatmap_data2, 
+
+
+ pheatmap::pheatmap(heatmap_data2,
  display_numbers = annotation_matrix,
  color = colorRampPalette(c("red", "white"))(100),
- cluster_rows = T, 
+ cluster_rows = T,
  cluster_cols = T,
  angle_col = 90,
  fontsize_row = 10,
  fontsize_col = 10,
  annotation_colors = list(display_numbers = c("black")),
  main = paste0("Heatmap of AIC Values for All Models","(",method,")")
  )
- 
+
 }