Full normalization assessment for simulated data — pip.simulated.data • precision.seq

Simulated data can be sampled using the function simulated.data.

pip.simulated.data(
  data,
  groups,
  norm.counts,
  adjust.factors,
  method.name,
  DE.method = "DE.voom",
  Pval = 0.01
)

Arguments

data	list of paired data sets. Each pair must consist of a benchmark data set "simulated_benchmark" and a test data set "simulated_test".
groups	Sample groups in the simulated test and benchmark data sets. Must be the same across all pairs.
norm.counts	Normalized counts of each simulated test data set for a user-provided normalization method.
adjust.factors	optional Adjustment factors for the normalization of the test data for a user-provided normalization method.
method.name	optional Name of the normalization method. Used for naming of data frames and plot legends.
DE.method	Method for computing differential expression statuses. Available: "DE.voom" and "DE.edgeR".
Pval	P-value cutoff for differential expression.

List of assessment statistics and plots (ggplot objects):

DE.stats: List of statistics (true positive rate, false positive rate, false discovery rate, and false negative rate) based on the comparison of differential expression statuses between the normalized test data and the unnormalized gold standard (the benchmark data). Contains results for each pair of simulated data sets.
p.boxplot: Boxplot for False Negative Rate (FNR) and False Discovery Rate (FDR) of the agreement of DE statuses between the simulated test and benchmark data sets.