GenOMICC_Influenza

Reproducing The GWAS

To run the GWAS, we only need the WDL-GWAS pipeline, make sure you have followed the installation isntructions and run the following:

input_config=config/genomicc-inputs.influenza.meta.plink2.json
compiled_input_config=config/genomicc-inputs.influenza.meta.plink2.dx.json
output_dir=/or3_influenza_plink2
commit=98c732469053100dff96cf3b2a4ebb9c70fafa43

# Clone WDL-GWAS pipeline
git clone https://github.com/olivierlabayle/WDL-GWAS
cd WDL-GWAS && git checkout $commit && cd ..

# Compile the workflow
java -jar $DX_COMPILER_PATH compile WDL-GWAS/workflows/gwas.wdl \
    -f -project $RAP_PROJECT_ID \
    -extras WDL-GWAS/config/extras.json \
    -reorg \
    -folder /workflows/gwas \
    -inputs ${input_config}

# Run the workflow
dx run -y \
    -f ${compiled_input_config} \
    --priority high \
    --preserve-job-outputs \
    --destination ${output_dir} \
    /workflows/gwas/gwas

Reproducing the Post-Analysis

Software Dependencies

• Install Julia, then install the package:

julia --project=. --startup-file=no -e 'using Pkg; Pkg.instantiate()'

• Install METAL.

Data Dependencies

• Per ancestry and trans-ancestry WDL-GWAS outputs downloaded from the RAP in $DATA_DIR/genomicc. The following files are expected:

Filename
data/genomicc/AFR.SEVERE_INFLUENZA.gwas.tsv
data/genomicc/EUR.SEVERE_INFLUENZA.gwas.tsv
data/genomicc/SAS.SEVERE_INFLUENZA.gwas.tsv

• FinGenn Data downloaded from the website.

Download the summary stats associated with:

• https://risteys.finngen.fi/endpoints/INFLUENZA
• https://risteys.finngen.fi/endpoints/J10_INFLUENZA

Move them to ./data/finngen and decompress with gunzip -d. Expected Files:

Filename
data/finngen/summary_stats_release_finngen_R12_INFLUENZA
data/finngen/summary_stats_release_finngen_R12_J10_INFLUENZA

Run The Meta-Analysis

julia --project=. --startup-file=no src/meta_analyse.jl data

Outputs will be in $DATA_DIR$/meta_analysis_workdir. The main output is the $DATA_DIR$/meta_analysis_workdir/META_ANALYSIS.all.tsv file.

Debugging a Run With A DNA Nexus Instance

To connect to a DNA Nexus instance, run:

instance_type=mem2_ssd1_v2_x8
dx run \
    --instance-type $instance_type \
    -imax_session_length="48h" \
    -y \
    --ssh app-cloud_workstation

You can download files via:

dx download file_id_1 file_id_2 ... 

And run the container with:

docker_tag=optional_fp
docker run -it --rm -v $PWD:/mnt/data olivierlabayle/wdl-gwas:$docker_tag /bin/bash

Then you can run code as normal, for instance to enter the Julia REPL:

julia --project=/opt/PopGen --startup-file=no --sysimage=/opt/PopGen/sysimage.so --threads=auto