Note:
I gave a similar talk at BGA last year. However, the paper has changed substantially in the meantime such that this year's presentation will be very different from what was presented last year. It would be good to update the BGA community on these developments, partially to avoid wrong applications or our method. In a nutshell, our method (GIV regression) focusses on controlling for pleiotropy in non-experimental, non-twin data. Pleiotropy is a potential source of bias in a wide range of scenarios which makes our approach potentially useful for many research questions. Furthermore, we demonstrate that our method can be used to estimate the chip heritability of a trait from polygenic scores.
Abstract:
Identifying causal effects in non-experimental data is an enduring
challenge. One proposed solution that recently gained popularity is
the idea to use genes as instrumental variables (i.e. Mendelian Randomization
– MR). However, this approach is problematic because
many variables of interest are genetically correlated, which implies
the possibility that many genes could affect both the exposure and
the outcome directly or via unobserved confounding factors. Thus,
pleiotropic effects of genes are themselves a source of bias in nonexperimental
data that would also undermine the ability of MR to
correct for endogeneity bias from non-genetic sources. Here, we
propose an alternative approach, GIV regression, that provides estimates
for the effect of an exposure on an outcome in the presence
of pleiotropy. As a valuable byproduct, GIV regression also provides
accurate estimates of the chip heritability of the outcome variable.
GIV regression uses polygenic scores (PGS) for the outcome of interest
which can be constructed from genome-wide association study
(GWAS) results. By splitting the GWAS sample for the outcome
into non-overlapping subsamples, we obtain multiple indicators of
the outcome PGS that can be used as instruments for each other,
and, in combination with other methods such as sibling fixed effects,
can address endogeneity bias from both pleiotropy and the
environment. In two empirical applications, we demonstrate that
our approach produces reasonable estimates of the chip heritability
of educational attainment (EA) and show that standard regression
and MR provide upwardly biased estimates of the effect of body
height on EA.
