Constructing GRNs from multimodal data?

Question

[kgoss]

Hello all,

I have a multimodal single cell dataset consisting of scRNA-seq and scATAC-seq. I'm wanting to construct gene regulatory networks, and was looking into SCENIC, however it seems that it only takes scRNA-seq as input. Does anyone know if I am able to use both modalities as input to SCENIC? Perhaps there is a way to get a weighted expression matrix from the Seurat WNN analysis? Any pointers are appreciated as this is my first time working with this kind of data.