Colony context and size-dependent compensation mechanisms give rise to variations in nuclear growth trajectories

The dev branch reflects the most up to date version of this analysis. To exactly reproduce the analysis as seen in bioRxiv manuscript use the bioRxiv-v1 branch.

The code in this repository generates all of the figures for Dixon et al. 2024 (bioRxiv). It is primarily intended to support reproducibility of our research. In addition, researchers may find parts of this code valuable for future work.

For a description of the cell treatments, imaging, and the purpose of each analysis, please refer to the paper.

The data used in this analysis are publicly available on Quilt under the Allen Insitute for Cell Science Terms of Use. The data are also available via the AWS S3 API directly in the folder s3://allencell/aics/nuc-morph-dataset.

Installation

Note

These are the basic installation steps. However, our recommendation is to install with pyenv and pdm. See advanced installation instructions here.

1. Install Python 3.9 and git. Update pip at least to 24.0.0.
2. Clone this git repository.

git clone [email protected]:AllenCell/nuc-morph-analysis.git
cd nuc-morph-analysis

3. Create a new virtual environment and activate it.

python -m venv venv
source venv/bin/activate

4. Install the required packages for your operating system. Replace linux with macos or windows as appropriate.

pip install -r requirements/linux/requirements.txt
pip install -e .

Reproduce figures

List available workflows with the following command.

python run_all_manuscript_workflows.py --list

Use the --only flag to run any one workflow. Confirm that your installation is working by running a fast workflow.

python run_all_manuscript_workflows.py --only error

This should write figures to the nuc-morph-analysis/nuc_morph_analysis/analyses/error_morflowgenesis/figures/ directory.

Important

Most workflows provided here are designed to run in a high-performance computing setting. They use 30-60GB of RAM and running all of them will take many hours even with a fast machine.

To run all the analyses in the paper, omit any options.

python run_all_manuscript_workflows.py

Figures are saved to the directories nuc-morph-analysis/nuc_morph_analysis/analyses/*/figures/.

Reproduce pre-processing of nuclear trajectory datasets

In addition to run_all_manuscript_workflows.py, this repository includes a few other entrypoints for specific pre-processing tasks.

• To reproduce the hiPSC single nuclei timelapse analysis datasets, run:

python nuc_morph_analysis/lib/preprocessing/save_datasets_for_quilt.py

• To reproduce 2024-09-19_baseline_intermediate_manifest.parquet, run:

python nuc_morph_analysis/lib/preprocessing/generate_main_manifest.py

• To reproduce the feeding control and inhibitor intermediate manifests, run:

python nuc_morph_analysis/lib/preprocessing/generate_perturbation_manifest.py

• To reproduce the timelapse feature explorer datasets, run:

python nuc_morph_analysis/lib/visualization/write_data_for_colorizer.py

Explore results in timelapse feature explorer

To interactively explore the results and features generated in this repository and shown in the paper, follow the links provided in this table.