create_ranking: Create a stability ranking for genes or miRNAs

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

Description

create_ranking function creates stability ranking of genes or miRNAs expression levels.

Usage

create_ranking(all_exp_data, all_uniq_samples, miRNA = FALSE)

Arguments

all_exp_data	List of expression matrices. The first element of the output of rep_elim function.
all_uniq_samples	List of character matrices with information about unique samples in each experiment. The second element of the output of rep_elim function.
miRNA	Logical value indicating if stability ranking should be performed for miRNA or genes. Default value is FALSE what means that ranking will be created for genes.

Details

Note that create_ranking works properly only if arguments are the output of rep_elim function so you should use it even if you know that there are no replications in your data. If there are no replications in the data rep_elim function will just prepared objects to use with create_ranking function.

The ranking is created with the following steps:

• Z-scores are calculated for each expression value for each gene within an experiment.

• For each treated sample a control is founded and logarithmic fold change calculated.

• For each gene within an experiment gene score is calculated as ratio between logarithmic control expression and logarithmic fold change.

• Stability index for each gene is calculated as a weighed mean of gene scores from all experiments.

Fold change in time for chosen genes or miRNAs could be plotted with lFC_in_time function. You can also generate summary for used experimental conditions with get_info_about_used_exp function.

It should be made into consideration that reliable ranking can be produced only with reasonable number of microarray data.

Value

Function returns a list. It's content depends on miRNA argument.

If miRNA == FALSE, the list contains as folows:

• $GeneRanking: stability ranking for genes

• $controls: pairs of treated and control samples

• $FC_data: matrices with fold changes

• $samples: names of treated samples

• $EntrezId: all Entrez Ids

else if miRNA == TRUE the list contains as folows:

• $miRNA_ranking: stability ranking for miRNAs

• $controls: pairs of treated and control samples

• $FC_data: matrices with fold changes

• $samples: names of treated samples

• $MIMATids: all MIMAT ids

See Also

rep_elim, lFC_in_time, get_info_about_used_exp

Examples

## Not run: 
##### Create stability ranking for genes

# 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("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)

# create ranking
gene_ranking = create_ranking(no_rep$noRepData, no_rep$uniqSamples, miRNA = FALSE)

##### Create stability ranking for miRNAs

# 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)

# create ranking
miRNA_ranking = create_ranking(no_rep$noRepData, no_rep$uniqSamples, miRNA = TRUE)

## End(Not run)

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