Comparing the Genetic Architecture of Lipid Traits between Populations

Comparing the Genetic Architecture of Lipid Traits between Populations

The majority of Genome-wide Association Studies, even now, focus on Caucasian, European populations. Genetic differences exist people of differing populations, where people of a certain ancestry or ethnicity can have genetic variants not present in those of European popualtions. These variants can lead to variable gene expression, as well as a differential response to drugs admistered. Hence, the resulsts of these single-poulation centric studies cannot be extrapolated to different populations. There is, therefore, an increased need to conduct such genome-wide studies in several other populations, along with accounting for population differences in large-scale studies.

I collected GWAS summary statistics of three lipid (cholesterol) biomarkers from seven different populations (British, two Greek isolates, Chinese, Japanese, East Asian, Ugandan) calculated the Polygeneic Risk Scores (PRSs) of each individual using the commonly overlapping lipid SNPs to assess if these scores varied across the populations. The heritability as well as trans-ancestral correlation was also calculated. The measured blood lipid biomarker levels were then examined across the genetic scores, to investigate the correlation between the blood lipid levels of one biomarker against the genetic score of another (biomarker-score cross association).

Software: UNIX, command-line R, PLINK, GEMMA, Popcorn

Results:

• The majority of European CVD/lipid loci overlap with the Japanese, Chinese, Greek-isolate, and African Ugandan populations.
• HDL biomarker/score showed an inverse relationship with LDL biomarker score; LDL and Triglycerides had a linear relationship, except for in the Ugandan population where triglyceride score did not correlate with biomarker score.
• The two Greek-isolate populations showed near perfect heritability and trans-ethnic correlation with the UK population, same as the Japanese and Chinese population with the East Asian population.

Project Highlights:

• Opportunity to work as as a Visiting Scientist at the Wellcome Sanger Institute
• Second-author publication in Nature Communications
• PRSs available on EMBL-EBI’s PGS Catalog
• First comparison study of its kind: Lipid phenotype (lipid blood levels) was found to be associated with lipid genotype (PRS)