Differential NicheNet code from Liver Atlas paper (Guilliams et al., Cell 2022)

In this repository you can find all the code that was used for the Differential NicheNet analyses and visualizations as described in the Liver Atlas paper from Guilliams et al.: Spatial proteogenomics reveals distinct and evolutionarily conserved hepatic macrophage niches.

Differential NicheNet is an extension of the default NicheNet pipeline to compare cell-cell interactions between different niches and better predict niche-specific ligand-receptor (L-R) pairs. It was used in that paper to predict ligand-receptor pairs specific for the Kupffer cell niche in mouse and human.

For users interested in applying Differential NicheNet to their own data, we recommend the vignette on the Github page of the nichenetr package Differential NicheNet analysis between niches of interest or Differential NicheNet analysis between conditions of interest.

Content of this repository

• scripts 0 and 1: scripts used to make Seurat objects from raw counts and cell annotations for the mouse data.
• scripts 2a, b and c: scripts used to annotate mouse stellate cells, LSECs and hepatocytes as portal vein or central vein cells.
• scripts 3a, b: scripts used to add this spatial zonation information to the existing Seurat object.
• scripts 4a, b, and c: scripts used to run the Differential NicheNet pipeline on the mouse data, comparing the Kupffer cell niche separately to respectively the bile duct macrophage niche, capsule macrophage niche and central vein macrophage niche.
• script 4d: script to analyze and combine these previous Differential NicheNet analyses, to come to a list of ligand-receptor pairs specific to the Kupffer cell niche (including code to make Figure S8G).
• script 5: script used to run the Differential NicheNet pipeline on the mouse data, comparing Kupffer cell niche to the niches of all the other liver macrophage niches combined.
• scripts 6 and 7: scripts used to make Seurat objects from raw counts and cell annotations for the human data.
• script 8: script used to run the Differential NicheNet pipeline on the human data, comparing Kupffer cell niche to the niches of all the other liver macrophage niches combined.
• script 9: script used to compare the mouse and human Differential NicheNet results, and look at evolutionary conservation, based on the output of scripts 5 and 8 (including code to make Figure 7B).
• script 10: script used to predict upstream ligands driving differences in KC-specific genes between human and mouse.
• script 11: script used to define and visualize the human-KC-specific ligand-receptor pairs, based on the output of scripts 9 and 10 (including code to make Figure 6E).

Notes:

• the exact data files you need to have in the data folder for running these scripts are available on this Zenodo page: https://zenodo.org/record/5840333. For exploration and downloading of other data from the paper, we refer to: Liver Atlas Data Portal

• the code provided in this repository is tailored to work on our gridengine cluster (via the qsub package) and will need to be adapted to work on your system/server/cluster.

References

Browaeys, R., Saelens, W. & Saeys, Y. NicheNet: modeling intercellular communication by linking ligands to target genes. Nat Methods (2019) doi:10.1038/s41592-019-0667-5

Guilliams et al. Spatial proteogenomics reveals distinct and evolutionarily conserved hepatic macrophage niches. Cell (2022) doi:10.1016/j.cell.2021.12.018