Partial Genetic Correlations between Traits

Abstract

Background Fibrosis involves excessive extracellular matrix deposition, disrupting tissue structure across multiple organs. While clinically heterogeneous, the extent to which fibrotic traits share genetic determinants remains unclear.

Methods This dissertation applied a multi-layered statistical genetics frame work to investigate shared genetic architecture across fibrosis-related traits. Genome-wide genetic correlation was first estimated using LD Score Regression. Partial genetic correlation then quantified the extent to which overlap could be explained by known mediators: body mass index (BMI), C-reactive protein (CRP) and alcohol consumption. Finally, Bayesian colocalisation was performed to assess whether overlapping signals reflected shared causal variants.

Results Genome-wide analyses revealed substantial overlap among fibrotic traits, particularly those involving metabolic and hepatic phenotypes. BMI accounted for a large proportion of this overlap; CRP contributed selectively. Alcohol consumption showed limited genome-wide sharing. Colocalisation identified the FTO locus as a shared signal between BMI and diabetes, but no colocalised signals were found for CRP or other fibrosis traits.

Conclusion These findings suggest that fibrosis reflects, in part, systemic genetic influences from metabolic and inflammatory pathways. While limited to European ancestry GWAS and common variants, the study provides new insight into shared aetiological mechanisms and potential targets for cross-organ intervention.