The ARIC model for MetaXcan is in Box:


The Atherosclerosis Risk in Communities Study (ARIC) generated genotype and proteomic data from a total of 9,084 participants (7,213 European Americans and 1,871 African Americans). The relative conectrations of plasma proteins or protein complexes was measured from blood samples using an aptamer-based approach. Genotyping of blood samples was imputed to the TOPMed reference panel (GRCh38).

Nilan Chatterjee et al analyzed cis-genetic regulation of the plasma proteome, generating PWAS through TWAS/Fusion pipeline. They study involved 4,665 SOMAmers measuring 4,491 unique plasma proteins or protein complexes encoded by 4,445 autosomal genes.

For details on the paper:

Generating the model

We created a prediction model compatible with MetaXcan software from the weights generated by Nilan Chatterjee et al’s PWAS study. The steps are documented below:

  1. Create prediction weights file
  2. Calculate covariances


We validated the model by running SPrediXcan on height and coronary artery disease GWAS, then comparing the results to association found from Whole Blood mashr models.

  1. GIANT height


Text and figures are licensed under Creative Commons Attribution CC BY 4.0. The source code is licensed under MIT.

Suggest changes

If you find any mistakes (including typos) or want to suggest changes, please feel free to edit the source file of this page on Github and create a pull request.


For attribution, please cite this work as

Sabrina Mi (2021). ARIC PWAS models. ImLab Notes. /post/2021/09/08/generating-metaxcan-prediction-model-from-aric-pwas/

BibTeX citation

  title = "ARIC PWAS models",
  author = "Sabrina Mi",
  year = "2021",
  journal = "ImLab Notes",
  note = "/post/2021/09/08/generating-metaxcan-prediction-model-from-aric-pwas/"