What is Mendelian Imputation Fgwas

Mendelian-imputation-fgwas is a workflow designed to perform family-based Genome-Wide Association Studies (GWAS) using the SNIPAR protocol, which incorporates imputations based on recent research.

Use cases

Use cases include analyzing the genetic basis of diseases, studying heritability of traits, and conducting research that requires the integration of familial data in GWAS.

How to use

To use mendelian-imputation-fgwas, define necessary variables in your terminal, create input files for FGWAS, and run the FGWAS model with the desired phenotype. Ensure to keep all variable definitions within the same terminal session.

Key features

Key features include the ability to run FGWAS with robust estimators, support for imputations, and the capability to handle various genetic data formats. It also allows for optional IBD segment inference.

Where to use

Mendelian-imputation-fgwas is primarily used in genetic research, particularly in studies related to complex traits and diseases, where understanding familial relationships is crucial.

Family-based GWAS workflow: SNIPAR

A protocol to run a FGWAS with imputations according to Young, et al. (2022); Guan, et al. (2025).

Run the workflow

1. Define the next variables in your console:

   export out_dir="/path/to/your/output/"
   export bed="/path/to/your/bedfile"
   export baseline="/path/to/your/baseline.csv"
   export kinship="/path/to/your/kinship.seg"
   export pcs="/path/to/your/pcs.txt"
   export chr_range="1 5 8-10 22" # string autosomal chromosomes to run
   

   Important

   Once you have defined the variables in your terminal, do not switch terminal. All the variables are defined within your terminal. Alternatively, you could define them in a .bashrc when creating your environment.

2. Create the input files for the FGWAS:

   python resources/build_inputs.py --baseline $baseline --kinship $kinship --pcs $pcs --outDir $out_dir
   

3. (Optional, and not recommended) Run the IBD segments inference. Just run it if you are doing tests, you are using special data or forgot to log it. It will take a while, so why not go for a coffee…

   ./run_ibd.sh
   

4. (Optional, and not recommended) Run the Mendelian Imputation. Same as above.

   ./run_imputation.sh
   

5. Run the FGWAS model you want to use. Provide a number for the phenotype you want to run the FGWAS on. For example, a FGWAS with the robust estimator (Guan et al., 2025) on BMI (125) would be:

   ./fgwas/robust.sh 125
   

Set up the environment

Manually

Useful for creating DNAnexus snapshots.

1. Install both versions of PLINK

   # PLINK 2.0
   wget https://s3.amazonaws.com/plink2-assets/alpha6/plink2_linux_avx2_20250420.zip && \
   unzip plink2_linux_avx2_20250420.zip -d plink2/ && \
   sudo mv plink2/plink* /usr/local/bin/ && \
   rm -rf plink2/ plink*.zip
   

   # PLINK 1.9
   wget https://s3.amazonaws.com/plink1-assets/plink_linux_x86_64_20241022.zip && \
   unzip plink_linux_x86_64_20241022.zip -d plink/
   sudo mv plink/plink /usr/local/bin/ && \
   rm -rf plink/ plink*.zip
   

2. Install Conda

   # Download CONDA
   wget https://repo.anaconda.com/miniconda/Miniconda3-latest-Linux-x86_64.sh
   

   # Run the installation script, and follow the instructions
   sh Miniconda3-latest-Linux-x86_64.sh
   

   # Activate it, and unable automatic start
   source ~/.bashrc && conda config --set auto_activate_base false
   

3. Install Rust

   # Download
   curl -sSf https://sh.rustup.rs | sh -s -- -y
   

   # Activate it
   . "$HOME/.cargo/env"
   

4. Install snipar

   conda create -n snipar_v0.0.22 python=3.9
   

   conda activate snipar_v0.0.22
   

   pip install snipar==0.0.22
   

Docker

A Dockerfile for the rolvhdez/snipflow:v0.1 image is provided. To build locally:

docker build -t snipflow .

Or you could also download the latest version in development.

docker pull rolvhdez/snipflow:latest