Error in `get_network_graph(object)`

[SharonKartika]

Hello. Thanks for the really great tool! I'm running it on a non model organism, following the getting started page and have run until the plot_module_metric(obj) function. I get an error at get_network_graph.

Computing gene-gene correlation
Computing weighted regulatory factor
Computing UMAP embedding

Warning in uwot(X = X, n_neighbors = n_neighbors, n_components = n_components,  :
  n_components > number of columns in input data: 2 > 0, this may give poor or unexpected results

Error in uwot(X = X, n_neighbors = n_neighbors, n_components = n_components, :
n_neighbors must be smaller than the dataset size

Looking at the source code for the get_network_graph function and stepping through manually, I think the issue is in this section:

    modules_use <- modules@meta %>%
        filter(target%in%colnames(rna_expr), tf%in%colnames(rna_expr))

modules_use in my case becomes a 0 element tibble.

What does it mean exactly for it to be 0? Is it that the data is not sufficient to infer GRN? Or is there some issue in the gene names or something like that?

Also, Running NetworkModules(obj) prints An Modules object with 22 TF modules. Is this sufficient information to construct the GRN on my own?

Thanks again for the tool.