DRomics (version 1.0-2)

omicdata: Import, check and normalization of omics data

Description

Data are imported from a .txt file, checked (see the description of argument file for the required format of data) and normalized if required.

Usage

omicdata(file, check = TRUE, norm.method = c("none", "cyclicloess", "quantile", "scale"))
# S3 method for omicdata
print(x, …)
# S3 method for omicdata
plot(x, …)

Arguments

file

The name of the .txt file containing one row per item, with the first column corresponding to the identifier of each item, in a column named "item", and the other columns giving the responses of the item for each replicate at each dose or concentration. The names of the corresponding columns must correspond to the tested doses or concentrations in a numeric format for the corresponding replicate (for example, if there are triplicates for each treatment, column names can be "item", 0, 0, 0, 0.1, 0.1, 0.1, etc.

check

If TRUE the format of the input file is checked.

norm.method

If "none" no normalization is performed, else a normalization is performed using the function normalizeBetweenArrays of the limma package using the specified method.

x

An object of class "omicdata".

…

further arguments passed to print or plot functions.

Value

omicdata returns an object of class "omicdata", a list with 5 components:

data

the numeric matrix of responses of each item in each replicate (one line per item, one column per replicate)

dose

the numeric vector of the tested doses or concentrations corresponding to each column of data

item

the character vector of the identifiers of the items, corresponding to each line of data

design

a table with the experimental design (tested doses and number of replicates for each dose) for control by the user

data.mean

the numeric matrix of mean responses of each item per dose (mean of the corresponding replicates) (one line per item, one column per unique value of the dose

norm.method

The normalization method specified in input

data.beforenorm

the numeric matrix of responses of each item in each replicate (one line per item, one column per replicate) before normalization

The print of a omicdata object gives the tested doses (or concentrations) and number of replicates for each dose, the number of items, the identifiers of the first 20 items (for check of good coding of data) and the normalization method. The plot of a omicdata object shows the data distribution for each dose or concentration and replicate before and after normalization.

Details

This function imports the data from a .txt file, then checks the format of data (see the description of argument file for the required format of data) and gives in the print information that should help the user to check that the coding of data is correct : the tested doses (or concentrations) the number of replicates for each dose, the number of items, the identifiers of the first 20 items. If the argument norm.method is not "none", data are normalized using the function normalizeBetweenArrays of the limma package using the specified method : "cyclicloess" (default choice), "quantile" or "scale".

References

Ritchie ME, Phipson B, Wu D, Hu Y, Law CW, Shi W, and Smyth, GK (2015), limma powers differential expression analyses for RNA-sequencing and microarray studies. Nucleic Acids Research 43, e47.

See Also

See normalizeBetweenArrays for details about the normalization.

Examples

# NOT RUN {
# (1) import, check and normalization of transcriptomics data 
# (an example on a subsample of a greater data set)
#
datatxt <- system.file("extdata", "transcripto_sample.txt", package="DRomics")

o <- omicdata(datatxt, check = TRUE, norm.method = "cyclicloess")
print(o)
plot(o)

# (2) normalization with other methods
(o.2 <- omicdata(datatxt, check = TRUE, norm.method = "quantile"))
(o.3 <- omicdata(datatxt, check = TRUE, norm.method = "scale"))

# }