TFPerturbation

This function enables in-silico transcription factor (TF) perturbation analysis using a regulatory network derived from co-expression and regulon information.

Usage

TFPerturbation(
  seurat_obj,
  selected_tf,
  pertub_dir,
  perturbation_name,
  graph = "RNA_nn",
  n_iters = 1,
  delta_scale = 1,
  corr_sigma = 0.05,
  n_threads = 4,
  use_velocyto = TRUE,
  use_graph_tp = FALSE,
  depth = 2,
  target_type = "both",
  use_regulons = TRUE,
  layer = "counts",
  slot = "counts",
  assay = "RNA",
  wgcna_name = NULL
)

Arguments

seurat_obj

A Seurat object.

selected_tf

The name of the transcription factor (TF) to perturb.

perturbation_name

A name for the in-silico perturbation that will be stored in the Seurat object.

graph

Name of the cell-cell graph in the Graphs(seurat_obj). Default = "RNA_nn".

n_iters

The number of times to apply the signal propagation throughout the TF regulatory network. Default = 1.

delta_scale

A numeric factor scaling the perturbation during propagation. Default = 1.

corr_sigma

A numeric scaling factor for the correlation matrix. Default = 0.05.

n_threads

Number of threads for the correlation calculation. Default = 4.

use_velocyto

Logical indicating whether to compute velocity-based transition probabilities. Default = TRUE.

use_graph_tp

Logical indicating whether to use the graph topology for transition probabilities. Default = FALSE.

depth

The depth of the regulatory network to use for target identification. Default = 2.

target_type

A string specifying the type of targets to include ("upstream", "downstream", or "both"). Default = "both".

use_regulons

Logical indicating whether to use regulons for TF-target relationships. Default = TRUE.

layer

Layer of the assay containing expression data. Default = "counts".

slot

Slot of the assay containing expression data. Default = "counts".

assay

Assay in seurat_obj containing expression information. Default = "RNA".

wgcna_name

The name of the hdWGCNA experiment in the seurat_obj@misc slot. If NULL, uses the active WGCNA experiment.

perturb_dir

A numeric determining the type of perturbation to apply. Negative values for knock-down, positive for knock-in, and 0 for knock-out.

Value

A Seurat object containing the in-silico TF perturbation results as a new assay.

Details

TFPerturbation enables in-silico transcription factor perturbation by simulating changes in TF activity and propagating these changes throughout the associated regulatory network. The workflow consists of the following steps:

1. Primary Perturbation: A primary in-silico perturbation is applied to the selected TF. The observed expression levels of the TF are adjusted by simulating changes using a perturbation direction (perturb_dir) and generating a new expression matrix.

2. Signal Propagation: The perturbation signal is propagated throughout the TF regulatory network using the adjacency matrix derived from TF-target relationships. The propagation process considers the network structure and propagates the signal over n_iters iterations.

3. Transition Probability Computation: Transition probabilities between cells are calculated based on the perturbation, optionally incorporating velocity or graph topology information.

This method leverages co-expression and regulatory network information to study the potential downstream effects of transcription factor perturbations on gene expression and cellular states.