The PRSmix package can be used to benchmark PRSs and compute the linear combination of trait-specific scores (PRSmix) and cross-trait scores (PRSmix+) to improve prediction accuracy.
NOTE:
- You can run all steps or run each of them individually (set a_runStep1 or a_runStep2 to true or false).
- Please make sure ALL optional FILES for all steps are provided even if you do not run that step (you can use a dummy file)
The pipeline contains 3 steps:
- Harmonizing SNP effects: s1_harmonize_SNPeffects
- Compute PRS: s2_computePRS
- Combine PRS: s3_combine_PRS
- This step makes sure all set of SNP weights aligned to the same risk allele. (Highly important)
- For more detail, please read more at PRSmix github part 1.
- Output: A file with Harmonized SNP effect.
- This step computes PRS with the SNP effects harmonized in step 1.
- For more detail, please read more at PRSmix github part 2.
- Output: A file with PRS computed
- This step performs PRS combination.
- For more detail, please read more at PRSmix github part 3.
- Output: A tarball with:
- The dataframe of training and testing sample split from the main dataframe:
_train_df.txtand_test_df.txt - The prediction accuracy in the training set for all PRSs:
_train_allPRS.txt - The prediction accuracy in the testing set for trait-specific PRSs:
_test_allPRS.txt - The prediction accuracy of PRSmix and PRSmix+:
_test_summary_traitPRS_withPRSmix.txtand_test_summary_traitPRS_withPRSmixPlus.txt
- The dataframe of training and testing sample split from the main dataframe:
Truong, B., Hull, L. E., Ruan, Y., Huang, Q. Q., Hornsby, W., Martin, H. C., van Heel, D. A., Wang, Y., Martin, A. R., Lee, H. and Natarajan, P. 2023. "Integrative Polygenic Risk Score Improves The Prediction Accuracy Of Complex Traits And Diseases". doi.org/10.1016/j.xgen.2024.100523.
This workspace is intended to be used for development of Consortium methods for PRSmix.
Please contact Buu Truong (btruong@broadinstitute.org) or Pradeep Natarajan (pradeep@broadinstitute.org) if you have any queries.