nichenet_seuratobj_aggregate_cluster_de Perform NicheNet analysis on Seurat object: explain differential expression (DE) between two 'receiver' cell clusters coming from different conditions, by ligands expressed by neighboring cells.
nichenet_seuratobj_aggregate_cluster_de(seurat_obj, receiver_affected, receiver_reference, condition_colname, condition_oi, condition_reference, sender = "all",ligand_target_matrix,lr_network,weighted_networks,expression_pct = 0.10, lfc_cutoff = 0.25, geneset = "DE", filter_top_ligands = TRUE, top_n_ligands = 20,top_n_targets = 200, cutoff_visualization = 0.33,organism = "human",verbose = TRUE)Single-cell expression dataset as Seurat v3 object https://satijalab.org/seurat/.
Name of cluster identity/identities of "affected" cells that were presumably affected by intercellular communication with other cells
Name of cluster identity/identities of "steady-state" cells, before they are affected by intercellular communication with other cells
Name of the column in the meta data dataframe that indicates which condition/sample cells were coming from.
Condition of interest in which receiver cells were presumably affected by other cells. Should be a name present in the "aggregate" column of the metadata.
The second condition (e.g. reference or steady-state condition). Should be a name present in the "aggregate" column of the metadata.
Determine the potential sender cells. Name of cluster identity/identities of cells that presumably affect expression in the receiver cell type. In case you want to look at all possible sender cell types in the data, you can give this argument the value "all". "all" indicates thus that all cell types in the dataset will be considered as possible sender cells. As final option, you could give this argument the value "undefined"."undefined" won't look at ligands expressed by sender cells, but at all ligands for which a corresponding receptor is expressed. This could be useful if the presumably active sender cell is not profiled. Default: "all".
The NicheNet ligand-target matrix denoting regulatory potential scores between ligands and targets (ligands in columns).
The ligand-receptor network (columns that should be present: $from, $to).
The NicheNet weighted networks denoting interactions and their weights/confidences in the ligand-signaling and gene regulatory network.
To determine ligands and receptors expressed by sender and receiver cells, we consider genes expressed if they are expressed in at least a specific fraction of cells of a cluster. This number indicates this fraction. Default: 0.10
Cutoff on log fold change in the wilcoxon differential expression test. Default: 0.25.
Indicate whether to consider all DE genes between condition 1 and 2 ("DE"), or only genes upregulated in condition 1 ("up"), or only genes downregulad in condition 1 ("down").
Indicate whether output tables for ligand-target and ligand-receptor networks should be done for a filtered set of top ligands (TRUE) or for all ligands (FALSE). Default: TRUE.
Indicate how many ligands should be extracted as top-ligands after ligand activity analysis. Only for these ligands, target genes and receptors will be returned. Default: 20.
To predict active, affected targets of the prioritized ligands, consider only DE genes if they also belong to the a priori top n ("top_n_targets") targets of a ligand. Default = 200.
Because almost no ligand-target scores have a regulatory potential score of 0, we clarify the heatmap visualization by giving the links with the lowest scores a score of 0. The cutoff_visualization paramter indicates this fraction of links that are given a score of zero. Default = 0.33.
Organism from which cells originate."human" (default) or "mouse".
Print out the current analysis stage. Default: TRUE.
A list with the following elements: $ligand_activities: data frame with output ligand activity analysis; $top_ligands: top_n ligands based on ligand activity; $top_targets: active, affected target genes of these ligands; $top_receptors: receptors of these ligands; $ligand_target_matrix: matrix indicating regulatory potential scores between active ligands and their predicted targets; $ligand_target_heatmap: heatmap of ligand-target regulatory potential; $ligand_target_df: data frame showing regulatory potential scores of predicted active ligand-target network; $ligand_activity_target_heatmap: heatmap showing both ligand activity scores and target genes of these top ligands; $ligand_receptor_matrix: matrix of ligand-receptor interactions; $ligand_receptor_heatmap: heatmap showing ligand-receptor interactions; $ligand_receptor_df: data frame of ligand-receptor interactions; $ligand_receptor_matrix_bonafide: ligand-receptor matrix, after filtering out interactions predicted by PPI; $ligand_receptor_heatmap_bonafide: heatmap of ligand-receptor interactions after filtering out interactions predicted by PPI; $ligand_receptor_df_bonafide: data frame of ligand-receptor interactions, after filtering out interactions predicted by PPI; geneset_oi: a vector containing the set of genes used as input for the ligand activity analysis; background_expressed_genes: the background of genes to which the geneset will be compared in the ligand activity analysis.
# NOT RUN {
seuratObj = readRDS(url("https://zenodo.org/record/3531889/files/seuratObj_test.rds"))
ligand_target_matrix = readRDS(url("https://zenodo.org/record/3260758/files/ligand_target_matrix.rds"))
lr_network = readRDS(url("https://zenodo.org/record/3260758/files/lr_network.rds"))
weighted_networks = readRDS(url("https://zenodo.org/record/3260758/files/weighted_networks.rds"))
nichenet_seuratobj_aggregate_cluster_de(seurat_obj = seuratObj, receiver_affected = "CD8 T", receiver_reference = "CD8 T", condition_colname = "aggregate", condition_oi = "LCMV", condition_reference = "SS", sender = "Mono", ligand_target_matrix = ligand_target_matrix, lr_network = lr_network, weighted_networks = weighted_networks)
# }
# NOT RUN {
# }
Run the code above in your browser using DataLab