NERVANALYTICAMissing RegulationGitHub ↗

About this project

Most schizophrenia GWAS loci are non-coding and presumed regulatory, yet only a minority colocalize with brain expression QTLs, a discrepancy known as the missing regulation problem[3]. This project quantifies the discrepancy on 281 PGC3 loci and tests six candidate explanations within one reproducible pipeline built entirely on public .

The headline findings: 32% of loci colocalize with at least one of 14 bulk brain expression datasets; eQTL sample size helps but saturates near 25% for single cortex datasets; fine-mapping reclassifies about a quarter of single-variant colocalization calls in each direction; single-cell, fetal, and splicing QTLs add a further 19 loci; and brain DNA methylation QTLs colocalize at 193 of 281 loci, including 118 of the 191 loci with no expression colocalization. Sixty-five loci (23%) remain unexplained by every channel. A transitive methylation-to-expression chain nominates candidate genes at 44 loci that lack direct eQTL support.

What we conclude, and what we do not

  • Adult bulk brain eQTLs anchor a minority of schizophrenia risk loci; the widely reported variant-to-gene gap replicates here at 32%.
  • Brain methylation QTLs colocalize at 69% of loci, so most risk variants without expression evidence are nonetheless active on brain regulatory DNA.
  • The shortfall therefore lies mostly with current expression catalogs, not with the regulatory hypothesis itself.
  • The methylation chain converts that observation into specific, testable gene nominations at 44 expression-orphan loci.
  • Colocalization demonstrates a shared causal variant, not causation; none of these results proves that a molecular trait transmits the disease effect.
  • Methylation links carry no direction or mechanism; mediation is one possibility among several.
  • Nominated genes are hypotheses for functional follow-up, not confirmed schizophrenia genes.
  • Unexplained loci may act in developmental windows, cell states, or modalities that public data do not yet cover.

Data sources

Methods in brief

Colocalization uses coloc.abf[5] with default priors over ±1 Mb windows around each index variant; multi-signal analysis uses SuSiE fine-mapping of the GWAS[15][16] paired with published per-gene SuSiE results via coloc.bf_bf[14]; the methylation chain scores each gene by the weaker link of GWAS↔CpG and CpG↔gene colocalization. The full Snakemake[9] pipeline, configuration, reports, and manuscript draft are in the repository; every number on this site regenerates from raw public downloads.

References

  1. Bryois J, et al. (2022). Cell-type-specific cis-eQTLs in eight human brain cell types identify novel risk genes for psychiatric and neurological disorders. Nat Neurosci. doi:10.1038/s41593-022-01128-z
  2. Byrska-Bishop M, et al. (1000 Genomes Project) (2022). High-coverage whole-genome sequencing of the expanded 1000 Genomes Project cohort including 602 trios. Cell. doi:10.1016/j.cell.2022.08.004
  3. Connally NJ, et al. (2022). The missing link between genetic association and regulatory function. eLife. doi:10.7554/eLife.74970
  4. de Klein N, et al. (2023). Brain expression quantitative trait locus and network analyses reveal downstream effects and putative drivers for brain-related diseases. Nat Genet. doi:10.1038/s41588-023-01300-6
  5. Giambartolomei C, et al. (2014). Bayesian test for colocalisation between pairs of genetic association studies using summary statistics. PLoS Genet. doi:10.1371/journal.pgen.1004383
  6. GTEx Consortium (2020). The GTEx Consortium atlas of genetic regulatory effects across human tissues. Science. doi:10.1126/science.aaz1776
  7. Hannon E, et al. (2016). Methylation QTLs in the developing brain and their enrichment in schizophrenia risk loci. Nat Neurosci. doi:10.1038/nn.4182
  8. Kerimov N, et al. (2021). A compendium of uniformly processed human gene expression and splicing quantitative trait loci. Nat Genet. doi:10.1038/s41588-021-00924-w
  9. Mölder F, et al. (2021). Sustainable data analysis with Snakemake. F1000Res. doi:10.12688/f1000research.29032.2
  10. Qi T, et al. (2018). Identifying gene targets for brain-related traits using transcriptomic and methylomic data from blood. Nat Commun. doi:10.1038/s41467-018-04558-1
  11. Roadmap Epigenomics Consortium (2015). Integrative analysis of 111 reference human epigenomes. Nature. doi:10.1038/nature14248
  12. Trubetskoy V, et al. (2022). Mapping genomic loci implicates genes and synaptic biology in schizophrenia. Nature. doi:10.1038/s41586-022-04434-5
  13. Walker RL, et al. (2019). Genetic control of expression and splicing in developing human brain informs disease mechanisms. Cell. doi:10.1016/j.cell.2019.09.021
  14. Wallace C (2021). A more accurate method for colocalisation analysis allowing for multiple causal variants. PLoS Genet. doi:10.1371/journal.pgen.1009440
  15. Wang G, et al. (2020). A simple new approach to variable selection in regression, with application to genetic fine mapping. J R Stat Soc B. doi:10.1111/rssb.12388
  16. Zou Y, et al. (2022). Fine-mapping from summary data with the “Sum of Single Effects” model. PLoS Genet. doi:10.1371/journal.pgen.1010299

github.com/Beyond-InFinnity/fine-mapping-and-missing-regulation ↗