TFActProfiler

TFActProfiler provides tools to infer transcription factor (TF) activities from gene expression data, assess their reliability, and simulate perturbation effects with or without additional model training.

Features

1. TF activity inference with reliability estimation

   • estimate_reliability: Computes activity scores for transcription factors and estimates their reliability

2. Perturbation simulation without additional training

   • perturbation_predict: Predicts the effects of TF perturbations (e.g., knockout) directly from prior TF–target interaction networks and observed expression data.
   • No extra model fitting required.

3. Perturbation simulation with additional training

   • train_W, predict_withW: Learns gene expression changes to predict TF perturbations.

Installation

pip install tfactprofiler

Quick usage

Usage examples are provided as Jupyter notebooks inside each example folder:

• example/TF_activity_inference.ipynb
  Single-cell TF activity inference with reliability estimation.

• example/Perturbation_without_training.ipynb
  Perturbation simulation without additional model training.

• example/Perturbation_with_training.ipynb
  Perturbation simulation with model training and cross-validation.