PRANA {PRANA}

PRANA

Description

A pseudo-value regression approach for differential network analysis that adjusts for additional covariates (PRANA).

Usage

PRANA(RNASeqdat, clindat, groupA, groupB)

Arguments

RNASeqdat

An RNA-Seq data with subjects in rows and genes in columns.

clindat

A data with clinical variables to be included in the regression (e.g., binary group variable indicating current smoking status, continuous age, ...)

groupA

Indices of the subjects in the first category (e.g., non-current smoker) of binary group variable.

groupB

Indices of the subjects in the second category (e.g., current smoker) of binary group variable.

Value

A list containing three data frame objects that summarize the results of PRANA. This includes beta coefficients, p-values, and adjusted p-values via the empirical Bayes approach for each predictor variables that are included in the regression model.

References

Ahn S, Grimes T, Datta S. A pseudo-value regression approach for differential network analysis of co-expression data. BMC Bioinformatics, 2023;24(1):8

Examples

data(combinedCOPDdat_RGO) # A complete data containing expression and clinical data.

# A gene expression data part of the downloaded data.
rnaseqdat = combinedCOPDdat_RGO[ , 8:ncol(combinedCOPDdat_RGO)]
rnaseqdat = as.data.frame(apply(rnaseqdat, 2, as.numeric))

# A clinical data with additional covariates sorted by current smoking groups:
# The first column is ID, so do not include.
phenodat = combinedCOPDdat_RGO[order(combinedCOPDdat_RGO$currentsmoking), 2:7]

# Indices of non-current smoker (namely Group A)
index_grpA = which(combinedCOPDdat_RGO$currentsmoking == 0)
# Indices of current smoker (namely Group B)
index_grpB = which(combinedCOPDdat_RGO$currentsmoking == 1)

PRANAres <- PRANA(RNASeqdat = rnaseqdat, clindat = phenodat,
groupA = index_grpA, groupB = index_grpB)