plot.PC.biplot

Source: R/plot.PC.biplot.R

Plot a biplot of the PCs from an object generated by perform.PCA. A biplot overlays sample scores (dots) with variable loadings (arrows), showing which proteins contribute most to the separation of samples along the selected principal components.

# S3 method for class 'PC.biplot'
plot(
  DEprot.PCA.object,
  PC.x = 2,
  PC.y = 1,
  color.column = "column.id",
  shape.column = NULL,
  label.column = NULL,
  plot.zero.line.x = TRUE,
  plot.zero.line.y = TRUE,
  n.loadings = 10,
  loading.color = "turquoise4",
  loading.arrow.size = 0.6,
  loading.label.size = 3,
  loading.alpha = 0.7,
  loading.scale = 0.8
)

Arguments

DEprot.PCA.object

An object of class DEprot.PCA, as generated by perform.PCA.

PC.x

Number indicating which Principal Component (PC) to display on the x-axis. Default: 2 (PC2).

PC.y

Number indicating which Principal Component (PC) to display on the y-axis. Default: 1 (PC1).

color.column

String indicating the name of the column in the metadata to use as factor for the dot colors. Default: "column.id" (each sample a color).

shape.column

String indicating the name of the column in the metadata to use as factor for the dot shapes. Default: NULL (all dots).

label.column

String indicating the name of the column in the metadata to use as label of the dots. Default: NULL (no labeling).

plot.zero.line.x

Logical value to indicate whether to plot two gray dashed line in correspondence of x=0. Default: TRUE.

plot.zero.line.y

Logical value to indicate whether to plot two gray dashed line in correspondence of y=0. Default: TRUE.

n.loadings

Integer indicating the number of top contributing protein loadings (arrows) to display on the biplot. Proteins are ranked by their Euclidean distance from the origin in the selected 2D PC plane. Default: 10.

loading.color

String indicating the color used for the loading arrows and labels. Default: "turquoise4".

loading.arrow.size

Numeric value indicating the linewidth of the loading arrows. Default: 0.6.

loading.label.size

Numeric value indicating the font size of the loading labels. Default: 3.

loading.alpha

Numeric value (0 to 1) indicating the transparency of the loading arrows and labels. Default: 0.7.

loading.scale

Numeric value used as a multiplier to manually adjust the scaling of the loading arrows relative to the sample scores. Values > 1 make arrows longer, values < 1 make them shorter. Default: 0.8.

Value

A ggplot object.

Details

The function extracts protein loadings from the PCA object stored in the prcomp slot. When the PCA was computed using stats::prcomp (no missing values), the protein projections from prcomp$x are used. When the PCA was computed using pcaMethods::pca (missing values present), the protein loadings from the loadings slot are used.
In both cases, the loadings are scaled so that the arrows fit visually within the range of the sample scores. Only the top n.loadings proteins (by Euclidean distance from origin in the selected PC plane) are displayed to keep the plot readable.
Gene/protein names are extracted from the row names of the loadings matrix.

Author

Sebastian Gregoricchio

Examples

# Perform Principal Component Analyses (PCA)
pca <- perform.PCA(DEprot.object = DEprot::test.toolbox$dpo.imp)

# Plot biplot of PC1-vs-PC2
plot.PC.biplot(DEprot.PCA.object = pca,
               color.column = "condition",
               shape.column = "replicate",
               n.loadings = 10)


# Plot biplot of PC2-vs-PC3 with more loadings
plot.PC.biplot(DEprot.PCA.object = pca,
               PC.x = 3,
               PC.y = 2,
               color.column = "condition",
               n.loadings = 15,
               loading.color = "gray30")