Question on Extending GenomicsSEM for Local Genetic Modeling

[yufeng huang]

Dear GenomicsSEM Team,

We have observed that the same phenotype may show different mixtures of genetic effects across distinct genomic regions, which can lead to underestimation of genetic correlations. To better capture fine-grained genetic relationships, we would like to ask whether genomicsSEM can be applied to multiple GWAS within the same local genomic region, or if the input must necessarily be genome-wide. Additionally, do you have any plans to extend genomicsSEM toward local genetic modeling approaches, similar to those implemented in tools like LAVA?

Best regards,

Yu-Feng Huang

[Elliot Tucker-Drob]

Stratified Genomic SEM is currently available, which allows for different multivariate models to be fit to different annotations (or the same model can be fit, and different parameter estimates obtained for each annotation). See links and papers below. However, annotations typically include many variants distributed widely across the genome, rather than a local region. We do not currently have an out-of-the-box means of applying different models specifically to individual LD blocks, or other spatially contiguous local regions of the genome. I would imagine that it would be straightforward to use the LAVA- or HESS- derived genetic covariance matrices as input. You would also need the full sampling covariance matrix (V) of the estimates (not just the SEs of the local h2s and cov_gs). I'm also not sure how stable the estimates are for these rather small regions given existing sample sizes- which is more of a concern when fitting SEMs as opposed to just determining whether a local rG is bonferroni-corrected significant.

https://github.com/GenomicSEM/GenomicSEM/wiki/6.-Stratified-Genomic-SEM

and the following papers:

Grotzinger, A. D., …Tucker-Drob, E. M., & Nivard, M. G. (2022). Genetic architecture of 11 major psychiatric disorders at biobehavioral, functional genomic, and molecular genetic levels of analysis. Nature Genetics [Tucker-Drob & Nivard jointly directed this work] Link

Grotzinger, A. D., de la Fuente, J., Davies, G., Nivard, M. G., & Tucker-Drob, E. M. (2022). Transcriptome-wide and stratified genomic structural equation modeling identify neurobiological pathways underlying general and specific cognitive functions. Nature Communications, 13, 1-15. Link

Grotzinger, A. D. et al. (2025). The Landscape of shared and divergent genetic influences across 14 psychiatric disorders. medRxiv. Link (to preprint)

--
You received this message because you are subscribed to the Google Groups "Genomic SEM Users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to genomic-sem-us...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/genomic-sem-users/853b2728-e5cc-4da8-bee5-99d7cc52248bn%40googlegroups.com.

[yufeng huang]

Thank you so much for the information. I think this will be a very interesting field.