Bioconductor EnhancedVolcano高亮指定基因的颜色设置问题求助

Hey there, let's work through fixing your volcano plot highlighting issue step by step:

Why Your Current Attempts Failed #

• Using the col parameter: The col argument in EnhancedVolcano is designed to take a 4-element vector, mapping to the four default gene groups: non-significant, FC-significant only, p-value significant only, and both FC/p-value significant. When you passed a vector matching your total gene count, the package ignored this input and defaulted to gray for all non-significant points.
• Using colCustom with errors: The "Aesthetics length" error almost always means your res_complete$picked_genes vector has missing values (NA) or doesn't exactly match the number of rows in res_complete. Double-check that length(res_complete$picked_genes) == nrow(res_complete) and there are no NAs in the logical vector.

Fixed Code #

Here's the corrected version that will properly highlight your target genes in forest green, with all other points in gray60:

# First, clean up any NA values in your picked_genes logical vector
res_complete$picked_genes[is.na(res_complete$picked_genes)] <- FALSE

# Generate a custom color vector matching every row in your results
custom_colors <- ifelse(res_complete$picked_genes, "forestgreen", "grey60")

# Build the volcano plot with corrected parameters
EnhancedVolcano(
  res_complete,
  lab = res_complete$gene_name,
  x = "log2FoldChange",
  y = "pvalue",
  pCutoff = 10e-3,
  FCcutoff = 1,
  xlim = c(-10, 10),
  ylim = c(0, -log10(10e-12)),
  colCustom = custom_colors,  # Use the pre-generated color vector
  pointSize = ifelse(res_complete$picked_genes, 5, 0.5),
  labSize = 2.5,
  selectLab = res_complete$gene_name[res_complete$picked_genes],  # Explicitly pass target gene names
  shape = 16,
  boxedLabels = TRUE,
  title = "DESeq2 results",
  subtitle = "Differential expression HC vs RA",
  caption = "FC cutoff, 1; p-value cutoff, 10e-3",
  legendPosition = "right",
  legendLabSize = 14,
  colAlpha = 0.9,
  drawConnectors = TRUE,
  hline = c(10e-8),
  widthConnectors = 0.2
)

Key Adjustments Explained #

• Removed conflicting shade parameters: The shade, shadeFill, and related arguments were competing with colCustom for control over point colors, so we removed them to avoid unexpected behavior.
• Explicit selectLab input: Instead of passing the logical vector directly, we extract the actual gene names from res_complete to ensure perfect matching with the lab parameter.
• NA cleanup: The line fixing NA values in picked_genes ensures your color vector has no missing entries, eliminating the aesthetics length error.

内容的提问来源于stack exchange，提问作者Malin