t1dgrs2

Generate a Type 1 Diabetes Genetic Risk Score (T1D GRS) that accounts for interactions between HLA DR-DQ risk haplotypes.

An improved T1D GRS (by Sharp et al., 2019) that incorporates both non-HLA and HLA risk components in discriminating between cases and controls, by accounting for interactions between HLA DR-DQ haplotype combinations. This uses 67 T1D-associated variants in either GRCh37 or GRCh38 to perform both a linear scoring of the genetic risk, with the added HLA DR-DQ interaction effect for the final GRS.

This Python package can be used on both imputed genotyping array and next generation sequencing (e.g., whole genome sequencing) input datasets, but will require quality control procedures to be applied beforehand.

Required: PLINK 1.9 fileset (.bed, .bim, .fam).

Please download all configuration files under the data directory from the Github page and adjust paths in t1dgrs2_setttings.yml accordingly.

Contributors

• Diane P Fraser (email)
• Seth A Sharp (email)
• Ankit M Arni (email)
• Richard A Oram (email)
• Michael N Weedon (email)
• Kashyap A Patel (email)

References

1. Oram RA, Patel K, Hill A, et al. (2016) A Type 1 Diabetes Genetic Risk Score Can Aid Discrimination Between Type 1 and Type 2 Diabetes in Young Adults. Diabetes Care 39(3): 337-344. 10.2337/dc15-1111.
2. Patel KA, Oram RA, Flanagan SE, et al. (2016) Type 1 Diabetes Genetic Risk Score: A Novel Tool to Discriminate Monogenic and Type 1 Diabetes. Diabetes 65(7): 2094-2099. 10.2337/db15-1690.
3. Sharp SA, Rich SS, Wood AR, et al. (2019) Development and Standardization of an Improved Type 1 Diabetes Genetic Risk Score for Use in Newborn Screening and Incident Diagnosis. Diabetes Care 42(2): 200-207. 10.2337/dc18-1785.

Requirements

• Linux environment (Debian-based or RHEL-based)
• The Linux x86_64 installations of any one of the following:
  • Miniforge
  • Mambaforge
  • An existing Anaconda installation (compatibility-based issues may occur if the conda package is very old, latest version would be preferable)
• PLINK v1.90 (minor version changes acceptable) command-line tool (bundled into the main package, no manual installation required for this)

Note: Please consult your local server admin to resolve conflicting installations of PLINK.

After setting up Anaconda, creating and activating a separate environment is highly recommended to avoid dependency conflicts with other packages already installed on the system, Python or otherwise.

conda env create -n <new_env_name> python
conda activate <new_env_name>

or the following, if Mambaforge was installed:

mamba env create -n <new_env_name> python
mamba activate <new_env_name>

Usage

After activating the environment created in the previous step, please install the t1dgrs2 package from the Bioconda repository using the following command:

conda install -n <new_env_name> -c bioconda t1dgrs2

or

mamba install -n <new_env_name> -c bioconda t1dgrs2

Once the package is installed, please use the -h flag to output the help text for more information:

python -m t1dgrs2 -h

The package has an built-in executable script, so simply run the module directly to generate the scores:

python -m t1dgrs2 -b /path/to/plink/bfiles/prefix -c /path/to/t1dgrs2_settings.yml -o /path/to/output/prefix

Execution

To highlight the flow of the program:

• A log file t1dgrs2.log will be created at your current working directory, to track execution progress in real-time.
• According to the value for argument -b / --bfile:
  • The input PLINK format files must be within the directory tree /path/to/plink/bfiles/.
  • The file names themselves must be of the form prefix.bed, prefix.bim and prefix.fam.
    • All three files must exist for successful execution.
• According to the value for argument -o / --output:
  • The final output files will be created within the directory tree /path/to/output/.
  • The file names themselves will be of the form prefix_FILE1, prefix_FILE2, etc.

If the -o argument is not set on execution, output files will be created as output_FILE1, output_FILE2, etc. at your current working directory.

Maintained by

Complex Traits Genomics of Diabetes group,
Department of Clinical and Biomedical Sciences,
University of Exeter Medical School,
St Luke's Campus, Heavitree Road,
Exeter, United Kingdom,
EX1 2LU.