15 January 2025 3 6K Report

Hi all!

I am currently working with the summary data provided by OneK1K for downstream post-GWAS analyses, such as SMR (Summary-data-based Mendelian Randomization) and two-sample Mendelian Randomization. However, I noticed that the dataset does not include the effect size (beta) for the eQTLs. Instead, it only provides a SPEARMANS_RHO value/column, which I assume represents some form of correlation.

I have a few questions regarding this issue:

  • Interpreting SPEARMANS_RHO: Can SPEARMANS_RHO be used as a proxy for the effect size (beta) in eQTL analysis? If not, is there a specific method to convert this metric into an effect size?
  • Data Preprocessing for SMR: The SMR tool requires effect sizes for each QTL. How should I preprocess the OneK1K data to convert it into the binary data format required by SMR? Specifically, what steps should I take to derive or estimate the effect sizes from the available SPEARMANS_RHO values?
  • Any insights are greatly appreciated!

    More Ao Lu's questions See All
    Similar questions and discussions