saeyslab / nichenetr

NicheNet: predict active ligand-target links between interacting cells
469 stars 117 forks source link

Issue Visualizing a Signaling Network #214

Closed HJamieJ closed 1 year ago

HJamieJ commented 1 year ago

Hi,

I tried to generate a signaling network tree with the data I am working with by following the instructions: https://github.com/saeyslab/nichenetr/blob/master/vignettes/ligand_target_signaling_path.md

Here is some of my code that could help you understand what is happening (after loading all the necessary RDS files as instructed):

ligands_BC_BC = c("FGF2", "BMP2", "AREG", "MSN") targets_BC_BC = subset(active_ligand_target_links_BC_to_BC_df, weight >= 0.1000) targets_BC_BC = unique(targets_BC_BC$target)

signaling_network_BC_BC = get_ligand_signaling_path(ligand_tf_matrix = ligand_tf_matrix, ligands_all = ligands_BC_BC, targets_all = targets_BC_BC, weighted_networks = weighted_networks)

signaling_network_BC_BC_min_max = signaling_network_BC_BC

signaling_network_BC_BC_min_max$sig = signaling_network_BC_BC_min_max$sig %>% mutate(weight = ((weight-min(weight))/(max(weight)-min(weight))) + 0.75) signaling_network_BC_BC_min_max$gr = signaling_network_BC_BC_min_max$gr %>% mutate(weight = ((weight-min(weight))/(max(weight)-min(weight))) + 0.75)

graph_min_max_BC_BC = diagrammer_format_signaling_graph(signaling_graph_list = signaling_network_BC_BC_min_max, ligands_all = ligands_BC_BC, targets_all = targets_BC_BC, sig_color = "indianred", gr_color = "steelblue") DiagrammeR::render_graph(graph_min_max_BC_BC, layout = "tree")

Resulted image attached.

I am speculating that it is caused by too many communications (4 ligands and 53 target genes), but I wanted to firstly ask if there was anything I needed to change in generating the code here. I believe the "active_ligand_target_links_BC_to_BC_df" is fine, considering that I had no problem generating the heatmaps following this: https://github.com/saeyslab/nichenetr/blob/master/vignettes/ligand_activity_geneset.md

Thanks!

Rplot

HJamieJ commented 1 year ago

It was the numbers of ligands and genes that were causing the issue.