This repository contains the code to reproduce the analysis in the following paper:
blisa: efficient mapping of cell–cell communication hotspots across spatial transcriptomics platforms
Lei Qin, Quoc Hoang Nguyen, Gordon K. Smyth, Lizhong Chen† and Yunshun Chen†
†These authors jointly supervised this work.
- Xenium and Visium sections of the same human breast cancer sample (Janesick et al., Nat Commun 2023, doi:10.1038/s41467-023-43458-x), as SpatialExperiment objects:
spe_xenium_bc_s1rep1.rdsandspe_visium_bc_s1.rds. The analysis scripts download them.
annotation/ contains:
janesick_2023_source_data.xlsx: source data of Janesick et al. (2023), with the pathology ROIs (Fig 4a) and the Visium deconvolution (Supp Fig 10);xenium_to_visium_transform.csv: the published Xenium-to-Visium image transform (10x Genomics);CellChatDB.human.rda: the CellChat human ligand–receptor database (jinworks/CellChat).
Detailed setup and analysis instructions are provided in the respective subfolders:
Analysis/: blisa on the Xenium and Visium data, and the benchmark of blisa, stLearn and SpatialDM.Figures/: figure panels regenerated from the analysis output.
blisa is available from CRAN and GitHub. The paper used blisa 1.0.2.
GPL-3; see LICENSE.