IntraTalkerPy

IntraTalkerPy combines two single-cell analysis workflows in one package:

• intratalkerpy.tf for transcription factor activity analysis and LR-to-TF workflows.
• intratalkerpy.perturbation for gene expression perturbation methods, utilities, and plotting.

📖 Documentation · Tutorials · API reference

Abstract

Single-cell sequencing has advanced the study of cell-cell communication, yet most methods focus on intercellular ligand-receptor interactions while neglecting downstream intracellular signalling cascades and the possibility that downstream target genes themselves encode ligands, thereby propagating communication across multiple cells. We present IntraTalker+CrossTalkeR that combines intracellular (IntraTalker) and intercellular (CrossTalkeR) signalling from multimodal single-cell data. IntraTalker infers cell-type-specific transcription factor activities and constructs receptomes that link receptors to downstream target genes, which are then integrated with ligand-receptor predictions in CrossTalkeR. To prioritize signalling receptors, the framework performs in silico receptor perturbation.

Overview

System requirements

• Operating system: tested on Ubuntu (Linux 6.8, x86_64)
• Python: 3.11–3.14, tested on 3.12.3.
• Hardware: normal desktop computer is sufficient for smaller data set and the Cell-Cell Communication analysis; consider using a HPC espcially for the perturbation simulations.
• Dependencies: installed automatically by pip (see pyproject.toml for the version ranges). The exact package versions every analysis was run with — Python and R — are listed under Session info.
• R is only needed for the IntraTalker/CrossTalkeR integration and the Sankey figures of the workflow, not for this package. See the setup page for that environment.

Installation

pip install git+https://github.com/CostaLab/IntraTalkerpy.git

or from a clone of this repository:

pip install .

git must be available on your PATH: one dependency (deltacorrpy) is installed from GitHub.

Typical install time on a normal desktop computer: under a minute (measured at ~40 s into a fresh virtual environment on Ubuntu with Python 3.12, downloading every dependency). Allow longer on a slow connection — the dependencies come to roughly 1 GB on disk.

Full instructions for both the Python and the R environment are on the setup page.

Demo

Use the two analysis areas as separate modules:

from intratalkerpy import tf
from intratalkerpy import perturbation

result = tf.IntraTalker_analysis(...)
perturbation.mt.simulation_of_perturbation(...)

A worked end-to-end example on real data, with the expected output of every step, is the Il1rn knockout bone marrow case study.

Expected run time on a normal desktop computer: most steps finish in seconds to a few minutes. The two long ones are both in step 9 — training the receptor to target gene model (~10 min) and simulating the receptor knockouts (~10 min).

Instructions for use

To run the workflow on your own data, follow the tutorials; each step lists the inputs it expects and the files it produces, and the API reference documents every function and its parameters.

To reproduce the results of the manuscript, run the Il1rn knockout bone marrow case study on the data deposited at Zenodo, with the package versions and random seeds listed under Session info.

Citation

If you use IntraTalkerPy, please cite:

Klöker V, Nagai JS, Feng Z, Mavrommatis L, Hermanns L, Moscoso JMJ, Ruiz M, Kuppe C, Costa IG. IntraTalker – Modelling Intracellular Signalling in Cellular Crosstalk. bioRxiv (2026). doi: 10.64898/2026.07.17.739174

License

MIT — see LICENSE.