An atlas connecting shared genetic architecture of human diseases and molecular phenotypes provides insight into COVID-19 susceptibility

Liuyang Wang, Thomas J. Balmat, Alejandro L. Antonia, Florica J. Constantine, Ricardo Henao, Thomas W. Burke, Andy Ingham, Micah T. McClain, Ephraim L. Tsalik, Emily R. Ko, Geoffrey S. Ginsburg, Mark R. DeLong, Xiling Shen, Christopher W. Woods, Elizabeth R. Hauser, Dennis C. KomedRxiv 2020 · DOI

Abstract

While genome-wide association studies (GWAS) have identified thousands of genetic variants linked to complex human traits, the biological mechanisms connecting these variants to disease pathophysiology often remain obscure. To bridge this gap, we developed the interactive Cross-Phenotype Analysis of GWAS database (iCPAGdb), a platform that leverages cross-phenotype associations to identify shared genetic architectures across clinical, cellular, and molecular scales. By incorporating linkage disequilibrium (LD) information, iCPAGdb accurately captures shared signals by proxy and calculates the significance of trait enrichment. Application of this atlas to severe COVID-19 revealed significant genetic overlap with idiopathic pulmonary fibrosis at the DPP9 locus and identified a potential regulatory mechanism at the ABO locus involving the SARS-CoV-2 receptor CD209 (DC-SIGN). These findings demonstrate how connecting genetically related traits across phenotypic scales can reveal novel biomarkers and therapeutic targets.

Introduction

Genome-wide association studies (GWAS) have been instrumental in mapping the genetic landscape of human health. However, a significant "mechanistic gap" persists: identifying a single nucleotide polymorphism (SNP) associated with a disease rarely explains how that variant contributes to the disease state. Understanding the path from genetic variation to clinical manifestation is essential for translating genomic findings into therapeutic interventions.

One powerful approach to bridging this gap is the study of pleiotropy—where a single genetic locus influences multiple traits. If a SNP is associated with both a clinical disease and a specific molecular trait (such as a protein level or metabolite concentration), that molecular trait may represent a functional link in the disease pathway. However, existing tools often struggle to account for the complex structure of the genome, specifically Linkage Disequilibrium (LD), or lack the computational scale to compare thousands of traits simultaneously.

The iCPAGdb Platform

To address these challenges, we introduced iCPAGdb (http://cpag.oit.duke.edu), an evolution of the Cross-Phenotype Analysis of GWAS (CPAG) tool. This platform provides a systematic framework for identifying shared genetic architecture across diverse biological scales.

Technical Innovations

  • LD Proxy Integration: Rather than relying solely on exact SNP overlaps, iCPAGdb utilizes ancestry-specific LD information from the 1000 Genomes Project. This allows the algorithm to identify shared signals where different SNPs within the same functional genomic block are associated with different traits.
  • Scalability and Speed: Utilizing an optimized SQL-based backend, the system can process user-uploaded GWAS summary statistics and compare them against a massive catalog of existing traits in under two minutes.
  • Statistical Rigor: The platform employs Fisher’s exact tests and the Benjamini-Hochberg procedure to ensure that identified overlaps are statistically significant, minimizing false discoveries in large-scale comparisons.

Multi-Scale Data Integration

iCPAGdb integrates a diverse array of datasets to create a comprehensive multi-scale atlas, including:

  • Clinical Traits: 3,793 traits from the NHGRI-EBI GWAS Catalog.
  • Cellular Phenotypes: 79 host-pathogen interaction traits from the H2P2 study.
  • Molecular Phenotypes: 491 blood metabolites and 55 urine metabolites.
  • COVID-19 Data: Recent GWAS data on severe respiratory failure and transcriptomic data from peripheral blood RNA-seq of 46 patients.

Insights into COVID-19 Pathophysiology

The utility of iCPAGdb was demonstrated through an analysis of severe COVID-19, which yielded two significant biological insights.

The DPP9 Locus and Lung Fibrosis

iCPAGdb identified a shared genetic signal between severe COVID-19 and idiopathic pulmonary fibrosis (IPF) at the DPP9 locus. Transcriptomic analysis of peripheral blood from COVID-19 patients further validated this link, showing that DPP9 expression is significantly induced in SARS-CoV-2 patients compared to healthy controls or those with bacterial infections. Furthermore, DPP9 levels were found to correlate with patient recovery, suggesting that the scarring observed in severe COVID-19 may share a common biological driver with chronic fibrotic lung diseases.

The ABO-CD209 Axis

While the association between the ABO blood group locus and COVID-19 risk is well-documented, the underlying mechanism has remained unclear. iCPAGdb revealed that the ABO locus colocalizes with the regulation of plasma protein levels of CD209 (DC-SIGN), a known receptor for SARS-CoV-2. Bayesian colocalization analysis (COLOC) confirmed a high probability (PP4 = 0.98) that the signals for COVID-19 and CD209 protein levels share a single causal variant. This suggests a model where ABO glycosyltransferase activity modifies CD209, potentially regulating viral entry and subsequent disease severity.

Results and Performance

The transition to iCPAGdb resulted in a 27.7-fold increase in direct cross-phenotype associations and a 47.7-fold increase in indirect (LD-proxy) associations compared to previous iterations. By transforming genomic coordinates into a list of potential biological pathways, the tool successfully bridges the gap between statistical association and biological mechanism.

Discussion and Future Directions

iCPAGdb provides a powerful resource for the research community to generate hypotheses regarding the pathophysiology of common diseases. However, the authors note that while the tool identifies significant associations, these do not constitute definitive proof of causation.

Future efforts will focus on:

  1. Functional Validation: Conducting "wet lab" experiments to test the proposed ABO-CD209 and DPP9 mechanisms in model systems.
  2. Expanding Data Resolution: Integrating emerging molecular GWAS data, such as single-cell eQTLs, to provide higher resolution into cell-type-specific mechanisms.
  3. Community Contribution: The continued utility of the atlas depends on the public availability of full GWAS summary statistics, highlighting the need for open data sharing in the genomics community.

By connecting clinical outcomes to molecular and cellular measurements, iCPAGdb offers a roadmap for identifying the next generation of biomarkers and therapeutic targets.

Summary updated

Abstract

While genome-wide associations studies (GWAS) have successfully elucidated the genetic architecture of complex human traits and diseases, understanding mechanisms that lead from genetic variation to pathophysiology remains an important challenge. Methods are needed to systematically bridge this crucial gap to facilitate experimental testing of hypotheses and translation to clinical utility. Here, we leveraged cross-phenotype associations to identify traits with shared genetic architecture, using linkage disequilibrium (LD) information to accurately capture shared SNPs by proxy, and calculate significance of enrichment. This shared genetic architecture was examined across differing biological scales through incorporating data from catalogs of clinical, cellular, and molecular GWAS. We have created an interactive web database (interactive Cross-Phenotype Analysis of GWAS database (iCPAGdb); http://cpag.oit.duke.edu) to facilitate exploration and allow rapid analysis of user-uploaded GWAS summary statistics. This database revealed well-known relationships among phenotypes, as well as the generation of novel hypotheses to explain the pathophysiology of common diseases. Application of iCPAGdb to a recent GWAS of severe COVID-19 demonstrated unexpected overlap of GWAS signals between COVID-19 and human diseases, including with idiopathic pulmonary fibrosis driven by the <i>DPP9</i> locus. Transcriptomics from peripheral blood of COVID-19 patients demonstrated that <i>DPP9</i> was induced in SARS-CoV-2 compared to healthy controls or those with bacterial infection. Further investigation of cross-phenotype SNPs with severe COVID-19 demonstrated colocalization of the GWAS signal of the <i>ABO</i> locus with plasma protein levels of a reported receptor of SARS-CoV-2, CD209 (DC-SIGN), pointing to a possible mechanism whereby glycosylation of CD209 by <i>ABO</i> may regulate COVID-19 disease severity. Thus, connecting genetically related traits across phenotypic scales links human diseases to molecular and cellular measurements that can reveal mechanisms and lead to novel biomarkers and therapeutic approaches.
BibTeX
@misc{wang2020atlas,
  author = {Liuyang Wang and Thomas J. Balmat and Alejandro L. Antonia and Florica J. Constantine and Ricardo Henao and Thomas W. Burke and Andy Ingham and Micah T. McClain and Ephraim L. Tsalik and Emily R. Ko and Geoffrey S. Ginsburg and Mark R. DeLong and Xiling Shen and Christopher W. Woods and Elizabeth R. Hauser and Dennis C. Ko},
  title = {{An atlas connecting shared genetic architecture of human diseases and molecular phenotypes provides insight into COVID-19 susceptibility}},
  year = {2020},
  doi = {10.1101/2020.12.20.20248572},
  howpublished = {medRxiv},
}
Figure