pre_process: Pre-Process the data

In espors/idepGolem: Integrated Differential Expression and Pathway Analysis

Pre-Process the data

Description

This function takes in user defined values to process the data for the EDA. Processing steps depend on data format, but generally includes missing value imputation, data filtering, and data transformations.

Usage

pre_process(
  data,
  missing_value = c("geneMedian", "treatAsZero", "geneMedianInGroup"),
  data_file_format = c(1, 2, 3),
  low_filter_fpkm,
  n_min_samples_fpkm,
  log_transform_fpkm,
  log_start_fpkm,
  min_counts,
  n_min_samples_count,
  counts_transform,
  counts_log_start,
  no_fdr
)

Arguments

data	Matrix of data that has already gone through convert_data()
missing_value	String indicating method to deal with missing data. This should be one of "geneMedian", "treatAsZero", or "geneMedianInGroup"
data_file_format	Integer indicating the data format. This should be one of 1 for read counts data, 2 for normalized expression, or 3 for fold changes and adjusted P-values
low_filter_fpkm	Integer for low count filter if data_file_format is normalized expression, NULL otherwise
n_min_samples_fpkm	Integer for minimum samples if data_file_format is normalized expression, NULL otherwise
log_transform_fpkm	TRUE/FALSE if a log transformation should be applied to normalized expression data
log_start_fpkm	Integer added to log transformation if data_file_format is normalized expression, NULL otherwise
min_counts	Numeric value for minimum count if data_file_format is read counts
n_min_samples_count	Integer for minimum libraries with min_counts if data_file_format is read counts
counts_transform	Integer to indicate which transformation to make if data_file_format is read counts. This should be one of 1 for log2(CPM+c) (EdgeR), 2 for variance stabilizing transformation (VST), or 3 for regulatized log (rlog)
counts_log_start	Integer added to log if counts_transform is log2(CPM + 2)
no_fdr	TRUE/FALSE to indicate fold-changes-only data with no p values if data_file_format is fold changes

Value

A list containing the transformed data, the mean kurtosis, the raw counts, a data type warning, the size of the original data, and p-values.

See Also

• cpm() for information on calculating counts per million

• vst() for information on variance stabilizing transformation

• rlog() for information on the regularized log transformation

Other preprocess functions: chr_counts_ggplot(), chr_normalized_ggplot(), eda_boxplot(), eda_density(), eda_scatter(), gene_counts_ggplot(), individual_plots(), mean_sd_plot(), rRNA_counts_ggplot(), total_counts_ggplot()

espors/idepGolem documentation built on Oct. 27, 2024, 4:56 a.m.