GWASDock

GWASDock is a collection of powerful genome-wide association study (GWAS) tools organized into reproducible Docker images. This project aims to simplify the reproduction of GWAS experiment, enable seamless transfer of GWAS tools across different computing environments, and improve productivity in GWAS analyses.

As an ongoing effort to enhance usability, the current release includes scripts that automate environment setup and container execution for each GWAS tool, as well as data format conversion between tools—surpassing the initial usability goals.

User Guides and Documentation

To facilitate user experience, we provide comprehensive, hands-on workflows for performing GWAS using GWASDock, our integrated multi-GWAS suite. These workflows include detailed instructions for data format conversion using scripts and for automating the launch of GWAS tools within Docker containers.

The documentation is organized as follows: documentation/examples/<corresponding GWAS tool folder>.

Specifically, the documentation for various GWAS tools is structured under documentation/examples/ with dedicated subfolders for each tool:

• PLINK
• TASSEL
• GAPIT
• GCTA
• FaST-LMM

Click on the links to access detailed workflows and instructions tailored for each tool.

Advanced Usage for Docker Experts

If you are familiar with Docker, you can manually build images from the provided Dockerfiles and mount the appropriate directories to run the containers. All related commands for building images and launching containers are organized within subfolders under documentation/usage/ for each GWAS tool.

The documentation provides detailed step-by-step instructions on how to reliably reproduce each container environment from scratch, including:

• Building the container image with docker buildx build or docker build using the specified Dockerfiles.
• Running each GWAS tool as a standalone container, either interactively or non-interactively, through the docker run commands.

Note: There are multiple ways to run these containers; users are encouraged to experiment to find the workflow that best fits their needs. For additional guidance, refer to the official Docker documentation: https://docs.docker.com/.

Tool-specific Documentation Links

Each tool has comprehensive instructions located in its subdirectory under documentation/usage/:

• PLINK Container:documentation/usage/plink-container
• TASSEL Container: documentation/usage/tassel-container
• GAPIT Container:documentation/usage/gapit-container
• GCTA Container:documentation/usage/gcta-container
  Note on Architecture: GCTA requires different Dockerfiles depending on whether your system architecture is linux/amd64 or linux/aarch64. The documentation details this distinction and provides the appropriate build and run commands. To identify your system architecture, run uname -a.
• FaST-LMM Container: documentation/usage/fastlmm-container

Currently Supported GWAS Tools

Since each GWAS tool comes with its own set of reference manuals, each GWAS tool's manual has been linked under its entry in the "currently supported GWAS tools" list that follows. You can use these manuals to better understand how to use each tool either interactively or non-interactively.

The currently supported GWAS tools are:

• PLINK (versions 1.9 and 2.0) by Chang et al. (2015) (DOI: 10.1186/s13742-015-0047-8)
  • The original PLINK (version 1.0) was developed by Purcell et al. (2007) (DOI: 10.1086/519795)
  • PLINK Usage Manual: The website that hosts the PLINK usage manual is located here: https://www.cog-genomics.org/plink/1.9/general_usage.
• TASSEL (version 5.2.95) by Bradbury et al. (2007) (DOI: 10.1093/bioinformatics/btm308)
  • Note on TASSEL: Only the command-line interface (CLI) for TASSEL is supported. The graphical user interface (GUI) version of TASSEL is less capable and stable than the CLI version, and it does not provide a scriptable interface to be used non-interactively as part of larger bioinformatics workflows. It is strongly recommended that you use the CLI version of TASSEL in any case, but it is nearly impossible to use the GUI version by design in the provided TASSEL container, since the user is expected to be familiar with the TASSEL CLI.
  • TASSEL Usage Manual: The website that hosts the TASSEL usage manual is located here: https://bitbucket.org/tasseladmin/tassel-5-source/wiki/UserManual.
• GAPIT (version 3) by Wang & Zhang (2021) (DOI: 10.1016/j.gpb.2021.08.005)
  • Version 2 of GAPIT was developed by Tang et al. (2016) (DOI: 10.3835/plantgenome2015.11.0120)
  • Version 1 (the original) of GAPIT was developed by Lipka et al. (2012) (DOI: 10.1093/bioinformatics/bts444)
  • GAPIT Usage Manual: The website that hosts the usage manual of GAPIT is located here: https://github.com/jiabowang/GAPIT/blob/master/Documents/gapit_help_document.pdf.
• GCTA (version 1.94.1) by Yang et al. (2011) (DOI: 10.1016/j.ajhg.2010.11.011)
  • GCTA Usage Manual: The website that hosts the GCTA usage manual is located here: https://yanglab.westlake.edu.cn/software/gcta/#Overview.
• FaST-LMM (version 0.6.12) by Lippert et al. (2011) (DOI: 10.1038/nmeth.1681)
  • GWASDock supports the following fork/implementation of FaST-LMM by Carl Kadie (Microsoft): FaST-LMM
  • FaST-LMM: The usage documentation for FaST-LMM is located here: https://fastlmm.github.io/FaST-LMM/.

System Requirements

A modest GNU/Linux computer is more than ample to use these tools. These recommended technical specifications are not exhaustive, and the user is always encouraged to get access to as much hardware horsepower they can get access to. Out of all the GWAS tools, GAPIT by Wang and Zhang (2021) is the most memory intensive, so the user is advised to use a GNU/Linux computer with access to more than 32 GiB of memory to be able to use GAPIT as it was designed/intended. Considering this, here are the recommended system specifications (hardware and software):

Hardware requirements:

• Central processing Unit (CPU): AMD Ryzen 5600 or Intel i5-10400 for x86 systems
• Memory: 32 GiB of random-access memory (RAM)
• Storage: Highly dependent on how much data you are working with, and how big the datasets are. The recommended minimum is 500 GiB of disk space, but this estimate can be easily insufficient if you are working with large datasets, and thus require more disk space to store them prior to running GWAS experiments.
• Graphical processing unit (GPU): Not necessary for this project, so use any GPU you want to.

Note: If you have a GNU/Linux aarch64 system – i.e., an arm computer, then you will need an arm computer with: (1) at-least 32 GiB of memory, (2) a reasonable amount of allocatable disk space, and (3) a modestly powerful arm CPU. Anything with at-least 4 cores, and good multithreaded and singlethreaded performance will serve GWAS experiments well. Most Amazon Web Services (AWS) and Microsoft Azure cloud servers provide performant arm instances with ample memory to easily handle the needs of these tools, so you can easily run the docker images of this project on cloud systems.

Software requirements:

• Any GNU/Linux distribution – for example: Arch Linux, Debian, Ubuntu, and/or Fedora Linux - that can run docker and containerd
• Docker client version: at-least v28.0
• Docker engine version: at-least v28.0
• containerd version: at-least v2.0
• runc version: at-least at-least v1.0
• Docker init version: at-least v0.15.0

References

Here is the references list:

• Bradbury, P. J., Zhang, Z., Kroon, D. E., Casstevens, T. M., Ramdoss, Y., & Buckler, E. S. (2007). TASSEL: Software for association mapping of complex traits in diverse samples. Bioinformatics, 23(19), 2633–2635. https://doi.org/10.1093/bioinformatics/btm308
• Chang, C. C., Chow, C. C., Tellier, L. C. A. M., Vattikuti, S., Purcell, S. M., & Lee, J. J. (2015). Second-generation PLINK: Rising to the challenge of larger and richer datasets. GigaScience, 4(1), 7. https://doi.org/10.1186/s13742-015-0047-8
• Lipka, A. E., Tian, F., Wang, Q., Peiffer, J., Li, M., Bradbury, P. J., Gore, M. A., Buckler, E. S., & Zhang, Z. (2012). GAPIT: Genome association and prediction integrated tool. Bioinformatics, 28(18), 2397–2399. https://doi.org/10.1093/bioinformatics/bts444
• Lippert, C., Listgarten, J., Liu, Y., Kadie, C. M., Davidson, R. I., & Heckerman, D. (2011). FaST linear mixed models for genome-wide association studies. Nature Methods, 8(10), 833–835. https://doi.org/10.1038/nmeth.1681
• Purcell, S., Neale, B., Todd-Brown, K., Thomas, L., Ferreira, M. A. R., Bender, D., Maller, J., Sklar, P., de Bakker, P. I. W., Daly, M. J., & Sham, P. C. (2007). PLINK: A tool set for whole-genome association and population-based linkage analyses. American Journal of Human Genetics, 81(3), 559–575. https://doi.org/10.1086/519795
• Tang, Y., Liu, X., Wang, J., Li, M., Wang, Q., Tian, F., Su, Z., Pan, Y., Liu, D., Lipka, A. E., Buckler, E. S., & Zhang, Z. (2016). GAPIT version 2: An enhanced integrated tool for genomic association and prediction. The Plant Genome, 9(2), 1–9. https://doi.org/10.3835/plantgenome2015.11.0120
• Wang, J., & Zhang, Z. (2021). GAPIT version 3: Boosting power and accuracy for genomic association and prediction. Genomics, Proteomics & Bioinformatics, 19(4), 629–640. https://doi.org/10.1016/j.gpb.2021.08.005
• Yang, J., Lee, S. H., Goddard, M. E., & Visscher, P. M. (2011). GCTA: A tool for genome-wide complex trait analysis. American Journal of Human Genetics, 88(1), 76–82. https://doi.org/10.1016/j.ajhg.2010.11.011