SCENIC+ analysis as a reference for new data?

[danieljrichard]

Hello!
I'm interested in using SCENIC+ to define GRNs for a reference set of multiome samples.
Given we have this Multiome data to learn regulatory relationships, I wondered if there was a way to use the GRNs to predict gene expression from new scATAC datasets. More accurately than the simplified techniques available in Seurat or ArchR, i.e. with some knowledge of the biological system in question. This seems to me to be an excellent use case for the learned GRNs, if possible