Combining SNP-to-gene linking strategies to identify disease genes and assess disease omnigenicity

Gazal, Steven; Weissbrod, Omer; Hormozdiari, Farhad; Dey, Kushal K.; Nasser, Joseph; Jagadeesh, Karthik A.; Weiner, Daniel J.; Shi, Huwenbo; Fulco, Charles P.; O’Connor, Luke J; Paşaniuc, Bogdan; Engreitz, Jesse M.; Price, Alkes L.

doi:10.1038/s41588-022-01087-y

Cited by 85 publications

(91 citation statements)

References 83 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We employed a widely used TWAS simulation framework (Mancuso Lab TWAS Simulator, see Code Availability ) to assess the power, bias, and calibration of TCSC in the presence of co-regulation across genes and tissues. We simulated a genome in which there are 1,245 genes, of which 125 (10%) are causal 34 . Each gene belongs to one of 249 independent genomic blocks, representing approximately 1Mb intervals of chromosome 1, which is 249 Mb in length.…”

Section: Methodsmentioning

confidence: 99%

“…The 10 tissues included one causal tissue explaining 5% of complex trait heritability and nine non-causal tissues. 10% of the genes (125 of the 1,245 genes) had nonzero (normally distributed) gene-trait effects in the causal tissue 34 . The SNP-heritability of the trait was set to 5%, all of which was explained by gene expression in the causal tissue; we note that TCSC is not impacted by SNP-heritability that is not explained by gene expression.…”

Section: Simulationsmentioning

confidence: 99%

See 1 more Smart Citation

Modeling tissue co-regulation to estimate tissue-specific contributions to disease

Amariuta

Siewert

Price

2022

Preprint

Self Cite

View full text Add to dashboard Cite

Integrative analyses of genome-wide association studies (GWAS) and gene expression data across diverse tissues and cell types have enabled the identification of putative disease-critical tissues. However, co-regulation of genetic effects on gene expression across tissues makes it difficult to distinguish biologically causal tissues from tagging tissues. While previous work emphasized the potential of accounting for tissue co-regulation, tissue-specific disease effects have not previously been formally modeled. Here, we introduce a new method, tissue co-regulation score regression (TCSC), that disentangles causal tissues from tagging tissues and partitions disease heritability (or covariance) into tissue-specific components. TCSC leverages gene-disease association statistics across tissues from transcriptome-wide association studies (TWAS), which implicate both causal and tagging genes and tissues. TCSC regresses TWAS chi-square statistics (or products of z-scores) on tissue co-regulation scores reflecting correlations of predicted gene expression across genes and tissues. In simulations, TCSC powerfully distinguishes causal tissues from tagging tissues while controlling type I error. We applied TCSC to GWAS summary statistics for 78 diseases and complex traits (average N = 302K) and gene expression prediction models for 48 GTEx tissues. TCSC identified 27 causal tissue-trait pairs at 10% FDR, including well-established findings, biologically plausible novel findings (e.g. aorta artery and glaucoma), and increased specificity of known tissue-trait associations (e.g. subcutaneous adipose, but not visceral adipose, and HDL). TCSC also identified 30 causal tissue-trait covariance pairs at 10% FDR. For the positive genetic covariance between eosinophil count and white blood cell count, whole blood contributed positive covariance while LCLs contributed negative covariance; this suggests that genetic covariance may reflect distinct tissue-specific contributions. Overall, TCSC is a powerful method for distinguishing causal tissues from tagging tissues, improving our understanding of disease and complex trait biology.

show abstract

Section: Methodsmentioning

confidence: 99%

Section: Simulationsmentioning

confidence: 99%

Modeling tissue co-regulation to estimate tissue-specific contributions to disease

Amariuta

Siewert

Price

2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…It should be further noted that ZBTB40 and SLC39A11 are the closest genes to their respective index SNPs. This is very relevant given that recent studies have shown that proximity to the index SNP has the best recall when predicting the causal gene within a GWAS locus, with precision and recall being the greatest within 1kb of the index SNP [ 53 , 54 ], which is the situation for SLC39A11 . While this is not definitive genetic proof of causality, these observations build a case for their causality, especially given their common link to calprotectin expression identified in the current study.…”

Section: Discussionmentioning

confidence: 99%

Prostaglandins and calprotectin are genetically and functionally linked to the Inflammatory Bowel Diseases

Karaky

Boucher²,

Mola³

et al. 2022

PLoS Genet

View full text Add to dashboard Cite

Background Genome wide association studies (GWAS) have identified and validated more than 200 genomic loci associated with the inflammatory bowel disease (IBD), although for most the causal gene remains unknown. Given the importance of myeloid cells in IBD pathogenesis, the current study aimed to uncover the role of genes within IBD genetic loci that are endogenously expressed in this cell lineage. Methods The open reading frames (ORF) of 42 genes from IBD-associated loci were expressed via lentiviral transfer in the THP-1 model of human monocytes and the impact of each of these on the cell’s transcriptome was analyzed using a RNA sequencing-based approach. We used a combination of genetic and pharmacologic approaches to validate our findings in the THP-1 line with further validation in human induced pluripotent stem cell (hiPSC)-derived-monocytes. Results This functional genomics screen provided evidence that genes in four IBD GWAS loci (PTGIR, ZBTB40, SLC39A11 and NFKB1) are involved in controlling S100A8 and S100A9 genes expression, which encode the two subunits of calprotectin (CP). We demonstrated that increasing PTGIR expression and/or stimulating PTGIR signaling resulted in increased CP expression in THP-1. This was further validated in hiPSC-derived monocytes. Conversely, knocking-down PTGIR endogenous expression and/or inhibiting PTGIR signaling led to decreased CP expression. These analyses were extended to the known IBD gene PTGER4, whereby its specific agonist also led to increased CP expression. Furthermore, we demonstrated that the PTGIR and PTGER4 mediated control of CP expression was dependent on signaling via adenylate cyclase and STAT3. Finally, we demonstrated that LPS-mediated increases in CP expression could be potentiated by agonists of PTGIR and PTGER4, and diminished by their antagonists. Conclusion Our results support a causal role for the PTGIR, PTGER4, ZBTB40, SLC39A11 and NFKB1 genes in IBD, with all five genes regulating the expression of CP in myeloid cells, as well as potential roles for the prostacyclin/prostaglandin biogenesis and signaling pathways in IBD susceptibility and pathogenesis.

show abstract

“…The current implementation of the MELD framework offers many directions for future development and applications. First, we note that in this study we did not incorporate additional variant annotations (e.g., based on epigenetic information, regulatory genomic units) during our computation of LD scores [62][63][64] .…”

Section: Discussionmentioning

confidence: 99%

Discovering non-additive heritability using additive GWAS summary statistics

Darnell

Smith

Ramachandran

et al. 2022

Preprint

View full text Add to dashboard Cite

The inflation of test statistics in genome-wide association (GWA) studies due to confounding factors such as cryptic relatedness, population stratification, and spurious non-zero genetic effects driven by linkage disequilibrium (LD) has been well characterized in the literature. The key theoretical contribution of this work is that epistasis (i.e., the interaction between multiple loci and/or genes) can also lead to bias in GWA summary statistics. To address this challenge, we develop marginal epistatic LD score regression and the accompanying software package MELD: an extended framework which takes in GWA test statistics and accurately partitions true additive genetic variation from non-additive genetic variation, as well as other biases. By re-analyzing 25 well-studied quantitative phenotypes from 349,468 individuals of European ancestry in the UK Biobank and up to 159,095 individuals in BioBank Japan, we illustrate that nonlinear effects are a significant source of signal in reported GWA summary statistics and provide evidence that epistasis is more widespread in human phenotypes than previously reported. Of the 25 complex traits we analyzed in the UK Biobank, 23 phenotypes have a significant amount of epistatic confounding captured within additive variation, including height, urate level, and cholesterol levels. The MELD software and its application to these biobanks represent a significant step towards resolving the true contribution of epistasis to human complex traits.

show abstract

Combining SNP-to-gene linking strategies to identify disease genes and assess disease omnigenicity

Cited by 85 publications

References 83 publications

Modeling tissue co-regulation to estimate tissue-specific contributions to disease

Modeling tissue co-regulation to estimate tissue-specific contributions to disease

Prostaglandins and calprotectin are genetically and functionally linked to the Inflammatory Bowel Diseases

Discovering non-additive heritability using additive GWAS summary statistics

Contact Info

Product

Resources

About