plot_pca: Plot PCA
In arnesmits/DEP: Differential Enrichment analysis of Proteomics data

Description Usage Arguments Value Examples

plot_pca generates a PCA plot using the top variable proteins.

1
2
3

plot_pca(dep, x = 1, y = 2, indicate = c("condition", "replicate"),
  label = FALSE, n = 500, point_size = 4, label_size = 3,
  plot = TRUE)

`dep`	SummarizedExperiment, Data object for which differentially enriched proteins are annotated (output from `test_diff()` and `add_rejections()`).
`x`	Integer(1), Sets the principle component to plot on the x-axis.
`y`	Integer(1), Sets the principle component to plot on the y-axis.
`indicate`	Character, Sets the color, shape and facet_wrap of the plot based on columns from the experimental design (colData).
`label`	Logical, Whether or not to add sample labels.
`n`	Integer(1), Sets the number of top variable proteins to consider.
`point_size`	Integer(1), Sets the size of the points.
`label_size`	Integer(1), Sets the size of the labels.
`plot`	Logical(1), If `TRUE` (default) the PCA plot is produced. Otherwise (if `FALSE`), the data which the PCA plot is based on are returned.

A scatter plot (generated by ggplot).

# Load example
data <- UbiLength
data <- data[data$Reverse != "+" & data$Potential.contaminant != "+",]
data_unique <- make_unique(data, "Gene.names", "Protein.IDs", delim = ";")

# Make SummarizedExperiment
columns <- grep("LFQ.", colnames(data_unique))
exp_design <- UbiLength_ExpDesign
se <- make_se(data_unique, columns, exp_design)

# Filter, normalize and impute missing values
filt <- filter_missval(se, thr = 0)
norm <- normalize_vsn(filt)
imputed <- impute(norm, fun = "MinProb", q = 0.01)

# Test for differentially expressed proteins
diff <- test_diff(imputed, "control", "Ctrl")
dep <- add_rejections(diff, alpha = 0.05, lfc = 1)

# Plot PCA
plot_pca(dep)
plot_pca(dep, indicate = "condition")