rep_elim: Elimination of replications within an experiment

In EwaMarek/FindReference: Find reference genes or miRNAs basing on microarray data.

Description

rep_elim function averages expression values for replications within an experiment and prepare expression matrices for being used in create_ranking function. It also returns information about unique samples in each experiment what is also needed for create_ranking function to run.

If there are no replications in the data rep_elim function will just prepared objects to use with create_ranking function.

Usage

rep_elim(processed_data, infoTables)

Arguments

processed_data	An expression matrix with normalized expression values or a list of them.
infoTables	A character matrix with information about experiment (or a list of them). The row names of it must be the same as column names in expression matrix (first argument) and the column names must be 'Experiment', 'Cell line', 'Treatment', 'Time', 'Dose' exactly in this order.

Details

Example table structure (second argument) should look like this:

	Experiment	Cell line	Treatment	Time	Dose
array1.txt	Exp1	HCT116	C	0	0
array2.txt	Exp1	HCT116	C	0	0
array3.txt	Exp1	HCT116	IR	1	4
array4.txt	Exp1	HCT116	IR	2	4
array5.txt	Exp1	HCT116	IR	4	2

Note that in treatment column it shoul be C for control arrays and IR for treated ones (even if the samples were treated with for example chemicals and not with ionizing radiation).

If you have downloaded data from ArrayExpress database you can find information needed for creating above in .sdrf files.

Value

Function returns list with two elements. The first one is an expression matrix without replications with changed column names. The second one is a character matrix with information about UNIQUE samples and is needed to create stability ranking.

See Also

create_ranking

Examples

## Not run: 
##### example with some experiments downloaded from ArrayExpress database

# download data from ArrayExpress database
to_download = c("E-GEOD-67309", "E-MTAB-966")
my_data = downloadAE(to_download, getwd())

# load data
platforms = c("Affymetrix", "Agilent")
loaded_data = load_multi_data(my_data, platforms)

# normalize and annotate
norm_data = multi_norm_and_annot(loaded_data$raw_expression_data, platforms)

# prepare tables for rep_elim function as shown in details
path_to_tables = system.file("inst/extdata", "tables_ex3.rds", package = "FindReference")
my_tables = readRDS(path_to_tables)

# eliminate replications and prepare object for create_ranking function
no_rep_data = rep_elim(norm_data, my_tables)


##### example with miRNA microarray data

# download data from ArrayExpress database
datamiRNA = downloadAE("E-MTAB-5197", "/home/emarek/")

# prepare table as shown in details load_miRNA help page
path_to_table = system.file("inst/extdata", "miRNA_ex1.rds", package = "FindReference")
my_table = readRDS(path_to_table)

# load data
loaded_data = load_miRNA(my_table, datamiRNA[[1]]$path)

# normalize and annotate data
norm_data = norm_and_annot_miRNA(loaded_data)

# eliminate replications and prepare object for create_ranking function
no_rep_data = rep_elim(norm_data, my_table)

## End(Not run)

EwaMarek/FindReference documentation built on May 30, 2019, 3:40 p.m.