Skip to main content

Conference Program

Subpage Hero

Loading

Use of Polygenic Risk Scores to Assess the Contribution of Common Genetic Variation in Primary Immunodeficiency Disorders

Laboratory Genetics and Genomics
  • Primary Categories:
    • Genomic Medicine
  • Secondary Categories:
    • Genomic Medicine
Introduction:
Primary immunodeficiency (PID) refers to a heterogeneous group of congenital disorders resulting from defects in the immune system. As PID patients can present with an extensive variability in symptoms, diagnosing the condition can be challenging. While genetic testing has identified numerous monogenic causes of PID, not all patients have one of these known variants. Furthermore, individuals with the same genetic variant can exhibit a broad spectrum of symptoms, complicating diagnostic efforts. Previous research suggests that these gaps in the understanding of PID aetiology might be explained by the cumulative effect of common variants. While common variants associated with PID have been identified, their utility in diagnosis or risk stratification through polygenic risk scores (PRS) has not been assessed. This study aims to clarify the role of common genetic variants in antibody deficiency-associated PID (AD-PID) and assess the utility of PRS in differentiating between healthy individuals and individuals with AD-PID.

Methods:
Genotypes of 8,298 individuals were analysed, including 7,672 unaffected controls and 626 AD-PID patients of European and European-Finnish ancestry. Two different size ratios of the training and testing samples were investigated, 80:20 and 60:40. To detect SNPs associated with AD-PID and determine their effect sizes, GWAS was performed using PLINK 1.9. The selection of SNPs for PRS calculation involved clumping and thresholding, utilizing parameters providing the best model fit, selected with the use of PRSice (v2.3.5). Following the creation of a model for calculation of AD-PID PRS in PLINK 1.9, its ability to differentiate between cases and controls was assessed and PRS comparisons were made among patients with a known monogenic cause of PID, carriers of PID-associated variants in TACI and individuals with an unknown genetic cause of PID. Furthermore, the relationship between AD-PID PRS and immunophenotype, with focus on B-cells, T-cells, natural killer cells and immunoglobulins, was evaluated.

 

Results:
Variants associated with AD-PID were identified, and optimal clumping and thresholding parameters for the polygenic risk score were determined. However, the resulting PRS model showed poor predictive accuracy, with a median R² of 0.009 in the 80:20 cohort split and 0.006 in the 60:40 split. The model did not demonstrate an ability to distinguish between AD-PID cases and controls, as indicated by a median AUC of 0.519 (80:20 cohort) and 0.516 (60:40 cohort). Additionally, there were no significant differences in PRS among individuals with distinct PID genotypes (P = 0.601 for monogenic cause vs. unknown genetic cause; P = 0.223 for monogenic cause vs. TACI variant). Finally, no significant correlation was found between PRS and immunophenotype, with Pearson correlation coefficients ranging from -0.11 to 0.25.



 

Conclusion:
The findings suggest that the identified variants do not sufficiently explain the differences between AD-PID cases and controls, nor do they account for variations among AD-PID patients with distinct genotypes or immunophenotypes. This may be due to the relatively small effect sizes of common variants, which are often subtle and require large sample sizes to detect meaningful associations. Additionally, the complex interplay between genetic and environmental factors in PID may dilute the effects of common variants, complicating their use in distinguishing clinical subgroups. These results indicate that larger studies with greater statistical power are likely required to more effectively assess the impact of common variants on AD-PID.

 

Agenda

Sponsors