Hundreds of cardiac MRI traits derived using 3D diffusion autoencoders share a common genetic architecture

The official code for the paper, "Hundreds of cardiac MRI traits derived using 3D diffusion autoencoders share a common genetic architecture". This repository contains all scripts used in this project, except for the deep learning pipeline (DL-pipeline). It includes scripts for pre-processing raw datasets from the UK Biobank, post-processing and analysing the latent embeddings, as well as conducting downstream analyses. The deep learning pipeline for unsupervised latent phenotyping using the 3D diffusion autoencoder can be found here: https://github.com/GlastonburyGroup/ImLatent.

This repository is organised into multiple folders, each containing standalone scripts for specific purposes. Additional scripts (e.g., for launching GWAS, performing post-GWAS analyses, PRS modelling, etc.) will be uploaded in the near future, along with detailed documentation. In the meantime, contact us for any questions or requests (Email: soumick.chatterjee@fht.org)

Execution

To simplify the installation of packages, Poetry can be used. Once Poetry is installed, this pipeline can be launched from its root directory without manually installing any dependencies manually by adding poetry run before calling Python. For example:

    poetry run python preprocess.createH5s.createH5_MR_DICOM.py

For continuous access in the terminal without adding the poetry run prefix to all commands, poetry shell (It must be installed additionally: https://github.com/python-poetry/poetry-plugin-shell) can be executed to activate the environment. The other Python commands can then be executed normally.

Citation

If you find this work useful or utilise any code from this repository in your research, please consider citing us:

@article{Ometto2024.11.04.24316700,
            author       = {Ometto, Sara and Chatterjee, Soumick and Vergani, Andrea Mario and Landini, Arianna and Sharapov, Sodbo and Giacopuzzi, Edoardo and Visconti, Alessia and Bianchi, Emanuele and Santonastaso, Federica and Soda, Emanuel M and Cisternino, Francesco and Pivato, Carlo Andrea and Ieva, Francesca and Di Angelantonio, Emanuele and Pirastu, Nicola and Glastonbury, Craig A},
            title        = {Hundreds of cardiac MRI traits derived using 3D diffusion autoencoders share a common genetic architecture},
            elocation-id = {2024.11.04.24316700},
            year         = {2024},
            doi          = {10.1101/2024.11.04.24316700},
            publisher    = {Cold Spring Harbor Laboratory Press},
            url          = {https://www.medrxiv.org/content/10.1101/2024.11.04.24316700},
            journal      = {medRxiv}
          }