Closed eleozzr closed 6 years ago
Hello,
Yes, all genes are used to infer the regulatory networks, but each gene can be calculated individually. Since it big datasets it can take several hours/days to run GENIE3 for all the genes, the execution is split into pieces to avoid the loss of the full run in case it is interrupted. This is not required, but rather a workaround for those users that cannot/do not want to use GRNBoost (the recommended method).
What's the function of parameter
nParts
inrunGenie3
function. In practice, how we specify thenParts
.SCENIC/R/runGenie3.R, line 30
genesSplit <- split(sort(rownames(exprMat)), 1:nParts)
and I don't understand why you usesort
? Shouldn't we put all genes together to infer gene regulatory networks from gene expression data??