This repository contains the code used to generate the figures in the paper "Spatial mapping of human hematopoiesis at single-cell resolution reveals aging-associated topographic remodeling".
The code is written in python and uses the following packages:
To install the packages, create new virtual environment and run the following command:
pip install -r requirements.txtThis will install the packages with the versions used in the paper.
Data to reproduce the figures is avaliable in associated Zenodo project data.
- data/NBM_dataset.h5ad - AnnData object containing the data used to generate the figures. To work with the dataset, download NBM_dataset.h5ad from Zenodo an put it into the data folder.
- data/clinical_annotation.csv - Clinical annotation for the samples in the AnnData object.
- data/platelets.xlsx - Platelet counts for the samples in the AnnData object.
- data/Values_for_Panel_1F_correlations.xlsx - Values for the correlations in Figure 1F.
Main AnnData contains graph neural network embeddings for community detection stored in X and mean expression data in obs with suffix '_mean_intensity'.
The notebooks used to generate the figures are in the notebooks folder. The notebooks are numbered in the order they should be run to reproduce the figures.