combineCOCOoutput: Combine COCONUT output from multiple objects into a single...

In COCONUT: COmbat CO-Normalization Using conTrols (COCONUT)

Description

Combine COCONUT output from multiple objects into a single object. Makes pooled analysis of COCONUT-co-normalized data easier.

Usage

combineCOCOoutput(COCONUT.out)

Arguments

COCONUT.out

Output from a call to COCONUT().

Details

The output from COCONUT() can be a bit daunting, and the separate dataobjects remain separated by input cohort, plus are separated into control and diseased components.

Value

This function will knit all data together into a list with three parts:

gene	contains a single matrix with all COCONUT-conormalized data (both control and disease)
pheno	contains a single data.frame with all phenotype info from the input samples, but ONLY from those columns whose colnames are same across all cohorts
class.cntl0.dis1	a binary vector that contains control/disease assignment for all columns in $genes.

Author(s)

Timothy E Sweeney, MD, PhD (tes17 [at] stanford [dot] edu)

References

Sweeney TE et al., "Robust classification of bacterial and viral infections via integrated host gene expression diagnostics", 2016

Examples

data(GSEs.test)

## apply COCONUT to a very small test case
## (3 datasets with 10 patients and 2000 genes)
GSEs.COCONUT <- COCONUT(GSEs=GSEs.test,
                        control.0.col="Healthy0.Sepsis1",
                        byPlatform=FALSE)

## combine output
GSEs.COCO.combined <- combineCOCOoutput(GSEs.COCONUT)
str(GSEs.COCO.combined)

COCONUT documentation built on May 2, 2019, 8:18 a.m.