Family-based GWAS have become the gold standard for assessing causal genetic effects. They’re often said to offer an unbiased estimate of the average causal effect (ATE) of an allele or PGS. Here, we (Carl Veller & @mollyprz.bsky.social ) evaluate such statements. www.biorxiv.org/content/10.1... 1/n - ThreadSky

About ThreadSky

gcbias.bsky.social • 530 days ago

Family-based GWAS have become the gold standard for assessing causal genetic effects. They’re often said to offer an unbiased estimate of the average causal effect (ATE) of an allele or PGS. Here, we (Carl Veller & @mollyprz.bsky.social ) evaluate such statements.
https://www.biorxiv.org/content/10.1101/2023.11.13.566950v1 1/n

Comments

gcbias.bsky.social•530 days ago

Claims about what can be learned from family-GWAS often rely on an analogy between the random transmission of alleles within families and natural experiments/RCTs, or are phrased in terms of counterfactual allele-swapping experiments. 2/n

gcbias.bsky.social•530 days ago

To make these analogies concrete, we develop a set of counterfactual definitions of the average phenotypic effects of allele & haplotype swaps in the presence of effect-size heterogeneity across individuals (GxE, GxG, or LD differences). 3/n

gcbias.bsky.social•530 days ago

We then consider whether family-based GWAS provide an unbiased estimate of average effects thus defined. We find that, under general conditions, family-GWAS cannot be claimed to estimate ATEs. 4/n

gcbias.bsky.social•530 days ago

The reason is that Mendelian segregation only randomizes alleles among children of heterozygotes, so the effects of alleles among the children of homozygotes are not observable in family-based GWAS. 5/n

gcbias.bsky.social•530 days ago

Therefore, if an allele has different average effects in the children of homozygotes and heterozygotes, as can arise from GxE, GxG, or diffs in LD, family-GWAS provide a biased estimate of the average effect in the population. 6/n

gcbias.bsky.social•530 days ago

At a single locus, family-GWAS can be reinterpreted as providing an unbiased estimate of the average effect in the children of heterozygotes—i.e., a local average treatment effect (LATE). However, this property does not extend to the effect of PGSs. 7/n

Posting Rules

Be respectful to others
No spam or self-promotion
Stay on topic
Follow Bluesky's terms of service

Comments

Posting Rules

Reply