Disclaimer: This article is for informational purposes only and is not intended to diagnose any conditions. LifeDNA does not provide diagnostic services for any conditions mentioned in this or any other article.

Overview

In 2021 a large research study was published on the genetics of various circulating biomarkers. The study focuses on evaluating the genetic basis of various blood and urine laboratory measurements , such as, lipids, glycemic traits, kidney function tests, and liver function tests, using data from the UK Biobank, which includes 363,228 individuals. The study aims to delineate the genetic basis of various circulating biomarkers, to understand their causal relationships with diseases, and to enhance the ability to predict genetic risk for common diseases.

 

Some highlights of the study are:

 

  1. Identification of Genetic Associations: The researchers identified 1,857 loci associated with at least one of the 35 traits they studied. These loci contained 3,374 fine-mapped associations and additional associations involving protein-altering gene variants, HLA (human leukocyte antigen) regions, and copy-number variants.
  2. Mendelian Randomization Analysis: The analysis revealed 51 causal relationships between the biomarkers and certain diseases. For example, they confirmed known effects such as the role of urate in gout and cystatin C in stroke.
  3. Polygenic Risk Scores (PRS): The study developed polygenic risk scores for each of the biomarkers and created ‘multi-PRS’ models that combined 35 PRSs simultaneously. These models improved genetic risk stratification for chronic kidney disease, type 2 diabetes, gout, and alcoholic cirrhosis in an independent dataset.

How Genetics Can Help Study Blood Markers

Serum and urine biomarkers are frequently measured to diagnose and monitor chronic disease conditions. Knowing the genetic predisposition to can have significant implications for disease treatment. The genetics of some biomarkers, such as lipids, glycemic traits, and kidney function measurements, have previously been extensively studied. However, in large population-scale datasets, the genetic basis of most biomarkers has not been explored.

 

To address this gap, the UK Biobank conducted laboratory testing of over 30 commonly measured biomarkers in serum and urine on a large cohort of more than 480,000 individuals, including both extensive phenotype and genome-wide genotype data. 

 

The present study systematically analyzed the genetic architecture and fine-mapped biomarker-associated loci in 363,228 individuals, including various genetic variants such as protein-altering, protein-truncating, non-coding, HLA, and copy number variants.  

 

Additionally, the researchers built phenome-wide associations for implicated genetic variants, evaluated causal relationships between biomarkers and 40 medically relevant phenotypes, and constructed polygenic prediction models. By understanding the genetic basis of these biomarkers, researchers can improve genetic risk stratification, enhance disease prediction models, and potentially develop better-targeted treatments for chronic diseases.

Breaking Down the Results

The researchers analyzed different genetic variants. This included directly genotyped and imputed autosomal variants, copy number variations (CNVs), and specific variants of genes within the human leukocyte antigen (HLA) system. The HLA system plays a crucial role in our body’s immune system. The study was done across 35 biomarkers in the UK Biobank. Participants were 318,953 White British, 23,582 non-British White, 6,019 African, 7,338 South Asian, and 1,082 East Asian individuals. The results from all groups, except the East Asian group, were combined for a meta-analysis with 355,891 individuals.

 

Researchers categorized the genetic variants into three groups: 

  • Protein-truncating
  • Protein-altering
  • Synonymous/non-coding variants. 

 

The findings agreed with previous studies on lipids, glycemic traits, kidney function tests, liver function tests, and other biomarkers. The researchers corrected the p-values for multiple testing and identified over 10,000 significant associations.

Heritability

Heritability estimates ranged from 0.6% for Lipoprotein A to 23.9% for IGF-1, from 3.2% for Microalbumin in urine to 57% for Total bilirubin. The researchers also assessed the polygenicity (effect of multiple genes on a single biomarker) of the biomarkers by calculating the fraction of total SNP heritability explained by the top 1% of SNPs. They found that three biomarkers—Lipoprotein A (67.7%), total bilirubin (60.9%), and direct bilirubin (57.5%)—had more than 50% of their SNP heritability explained by the top 1% of loci. 

 

The remaining 32 biomarkers exhibited moderate to high polygenicity. This finding is  vital because it helps determine how many genetic variants contribute to the heritability of each trait. This knowledge can guide the development of more accurate genetic risk prediction models, inform the design of future studies, and enhance our understanding of the underlying biological mechanisms influencing these biomarkers.

Therapeutic Targets

The study found 58 protein-truncating variants and 1,323 protein-altering variants outside the MHC region that were significantly impacting biomarker levels. By analyzing 166 traits in the UK Biobank the study identified 57 phenotype associations, including novel discoveries. For cardiovascular biomarkers, key genetic variants were linked to cholesterol, triglycerides, and heart disease. Liver biomarkers revealed variants affecting enzyme levels and gallstone risk. Kidney biomarkers showed associations with kidney disease and function. Bone, joint, glucose, HbA1C, and hormone biomarkers highlighted genetic variants with various effects, such as diabetes risk and growth factor levels. These results suggest that understanding the genetic basis of biomarker levels can help identify disease mechanisms and potential therapeutic targets.

What Does This Mean for Us?

This study is valuable as it enhances our understanding on how genetic variations influence biomarker levels and their association with disease phenotypes. By identifying protein-altering variants and their potential therapeutic targets, the study provides valuable insights for developing new treatments. Furthermore, the predictive models combining polygenic risk scores with biomarker data improve the accuracy of disease outcome predictions. This can be particularly beneficial for risk stratification and personalized medicine. The combined resource of association summary statistics, fine-mapped regions, and polygenic prediction models that the study provides, can inform future research and clinical practices, aiding in better disease prevention and management strategies.

References

  1. https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6786975/