Calculate the difference between the clock correlation distances (CCDs), relative to a reference, for two groups of samples. Statistical significance is calculated using permutation of the samples that belong to either of those two groups.
calcDeltaCCD( refCor, emat, groupVec, groupNormal, refEmat = NULL, nPerm = 1000, geneNames = NULL, dopar = FALSE, scale = FALSE )
| refCor | Correlation matrix to be used as the reference, such as comes
from |
|---|---|
| emat | Matrix of expression values, where each row corresponds to a gene
and each column corresponds to a sample. The rownames and colnames of
|
| groupVec | Vector indicating the group to which group each sample belongs. It's ok for groupVec to have more than two groups. |
| groupNormal | Value indicating the group in groupVec that corresponds to normal or healthy. Other groups will be compared to this group. |
| refEmat | Optional expression matrix for calculating co-expression for
the reference, with the same organization as |
| nPerm | Number of permutations for assessing statistical significance. |
| geneNames | Optional vector indicating a subset of genes in |
| dopar | Logical indicating whether to process features in parallel. Make sure to register a parallel backend first. |
| scale | Logical indicating whether to use scaled CCDs to calculate difference. |
A data.table with columns for group 1, group 2, deltaCCD, and
p-value. In each row, the deltaCCD is the CCD of group 2 minus the CCD of
group 1, so group 1 corresponds to groupNormal.
if (FALSE) { library('deltaccd') library('doParallel') library('doRNG') registerDoParallel(cores = 2) set.seed(35813) refCor = getRefCor() ccdResult = calcCCD(refCor, GSE19188$emat, GSE19188$groupVec, dopar = TRUE) deltaCcdResult = calcDeltaCCD( refCor, GSE19188$emat, GSE19188$groupVec, 'non-tumor', dopar = TRUE) pRef = plotRefHeatmap(refCor) pTest = plotHeatmap(rownames(refCor), GSE19188$emat, GSE19188$groupVec) }