paperKB
coga / coga-kb
Processing
Help
Sign in

Chunk #52 — Discussion — Limitations of the proposed approach

Source
Mendelian randomization with invalid instruments: effect estimation and bias detection through Egger regression.
Embedded
yes

Text

In this article, we have assumed that genetic variants are uncorrelated. If the variants are correlated, and the correlations between variants are known, they can therefore be used within generalized weighted linear regression instead of weighted linear regression in either the IVW or MR-Egger method, incorporating the correlations into the weighting matrix. Further work is currently being undertaken to explore this method.