deTS (version 1.0)

tsea.expression.normalization: RNA-Seq expression profiles normalization

Description

To avoid the data bias and adapt better data heterogeneity, before tsea.expression.decode() analysis, the raw discrete RPKM value have to normalized to continuous variable meet the normal distribution before t-test.

Usage

tsea.expression.normalization(query_mat, 
correction_factor, normalization = "abundance")

Arguments

query_mat

a RNA-seq RPKM object, row name should be gene symbol, and column name should be sample name.

correction_factor

correction_factor, a gene table object contain genes average expression level and standard variance in GTEx database, can be loaded by data(correction_factor).

normalization

normalization methods, c("z-score", "abundance")

Value

A data frame with normalized RNA-Seq expression profiles.

Rows stand for tissue names and columns stand for sample names.

Details

As RNA-Seq samples are often heterogeneous, before in-depth analysis, it is necessary to decode tissue heterogeneity to avoid samples with confounding effects. However, the raw discrete RPKM value have to normalized to continuous variable meet the normal distribution before t-test.

References

Pei G., Dai Y., Zhao Z., Jia P. (2019) deTS: Tissue-Specific Enrichment Analysis to decode tissue specificity. Bioinformatics, In submission.

See Also

https://github.com/bsml320/deTS

Examples

# NOT RUN {
data(query_GTEx)
query_matrix = query_GTEx[,1:2]
data(correction_factor)
query_mat_zscore_nor = tsea.expression.normalization(query_matrix, 
	correction_factor, normalization = "z-score")
# }