BSH

This is the code for block-shaped heterogeneity as an emergent organizing principle of spatial heterogeneity across human solid tumors.

Directory structure

The code for BSH is organized into different directories and scripts.

The directory structure is as follows:

• bin :  This directory contains ggplot2 theme and function that is needed for regenerating plots.
• data : This directory contains data that is needed for regenerating plots and descriptions for the data .
• figure1-5 : This directory contains the analysis scripts for regenerating main figures .
• simulation :  This directory contains the deme-based simulation scripts and descriptions for the usage .

Description of Scripts and How to Use

figure1-5

• Description: These folders contain R scripts to generate main plots in the article. All data can be found in the data directory. You can reproduce the results by running these R scripts like Rscript figure1.R .

simulation

Description: This folder contains python scripts to simulate deme-based tumor. The simulation parameters and instructions for use can be found in the README file in this directory.

Dependencies

R:

### R Dependencies
R version 4.3.1 (2023-06-16)
Platform: aarch64-apple-darwin20 (64-bit)
Running under: macOS Sonoma 14.6.1
|Package          |Version |
|:----------------|:-------|
|adegenet         |2.1.10  |
|ape              |5.7.1   |
|car              |3.1.2   |
|cowplot          |1.1.3   |
|GET              |1.0.4   |
|ggplot2          |3.5.2   |
|ggpmisc          |0.5.5   |
|ggpubr           |0.6.0   |
|ggtext           |0.1.2   |
|ggtree           |3.10.1  |
|graphics         |4.3.1   |
|ks               |1.14.2  |
|landscapemetrics |2.2.1   |
|mobster          |1.0.0   |
|patchwork        |1.2.0   |
|phangorn         |2.11.1  |
|purrr            |1.0.2   |
|raster           |3.6.26  |
|spatstat         |3.3.0   |
|spdep            |1.3.13  |
|terra            |1.7.78  |
|this.path        |2.4.0   |
|tidyverse        |2.0.0   |
|glmmTMB          |1.1.11  |
|IOBR             |2.2.2   |

Python：

### Python Dependencies for simulation
Python version 3.x
- numpy
- sys
- math
- random
- heapq
- subprocess
- collections

Data source

The dataset comprises 165 patients, including private sequencing data generated for this study (five HCC patients, DT42–DT51, and eight LUAD patients). The remaining 152 patients were collected from previously published cohorts, as listed in /data/fig1_patientID_clean.csv. For all 165 patients, VAF matrices are stored in /data/vaf_public_liver_lung_final, maximum parsimony trees are stored in /data/rds_public_liver_lung_final, and two-dimensional coordinate files are stored in /data/fig2_public_liver_lung_location_absolute_nsr_1169_final.csv.

License

The source code in this repository is licensed under the GNU General Public License v3.0. See the LICENSE.txt file.

Original data, metadata, documentation, and other non-software materials produced for this study are licensed under the Creative Commons Attribution 4.0 International License. See LICENSE-DATA.txt.

Data obtained or derived from previously published cohorts remain subject to the licensing and reuse conditions of their original sources.