CrossTalker

James S. Nagai¹, Nils B. Leimkühler², Rebekka K. Schneider^2,3,4, Ivan G. Costa^1*

¹Institute for Computational Genomics, Faculty of Medicine, RWTH Aachen University, Aachen, 52074 Germany

²Department of Hematology, Erasmus Medical Center, Rotterdam, 3015GD, the Netherlands

³Department of Cell Biology, Institute for Biomedical Engineering, Faculty of Medicine,RWTH Aachen University, Pauwelsstrasse 30, 52074 Aachen, NRW, Germany

⁴Oncode Institute, Erasmus Medical Center, Rotterdam, 3015GD, the Netherlands

Motivation: Ligand-receptor (LR) analysis allows the characterization of cellular crosstalk from single cell RNA-seq data. However, current LR methods provide limited approaches for prioritization of cell types, ligands or receptors or characterizing changes in crosstalk between two biological conditions.

Results: CrossTalkeR is a framework for network analysis and visualisation of LR networks. CrossTalkeR identifies relevant ligands, receptors and cell types contributing to changes in cell communication when contrasting two biological states: disease vs. homeostasis. A case study on scRNA-seq of human myeloproliferative neoplasms reinforces the strengths of CrossTalkeR for characterisation of changes in cellular crosstalk in disease state.

Install

You can install CrossTalkeR with the simple comands below:

install.packages("devtools")
devtools::install_github("https://github.com/CostaLab/CrossTalkeR")
require(CrossTalkeR)

CrossTalkeR Plots examples and vignette

We provide in our vignette examples on how to analyse cell interactions from a human myelofibrosis single cell RNA-seq.

vignette('CrossTalkeR')

Integration of CellPhoneDB with CrossTalkeR

We provide here a tutorial on how to integrate preditions from CellPhoneDB into Crosstalker.

CellPhoneDB Tutorial.md

Current Features v1.0.0

• Single and Comparative Reports
  • Cell Cell Interaction visualization
  • Sending and Receiving Cells Ranking
  • Gene Target based Sankey Plots
  • CCI and GCI PCA ranking
    • All measures and PC table
    • PC1 and PC2 based barplot
  • Leimkühler et. al. [2] data were added to the package

References

[1] CrossTalkNet: Analysis and Visualisation of LigandReceptor Networks. (in preparation)

[2]Heterogeneous bone-marrow stromal progenitors drive myelofibrosis via a druggable alarmin axis. link