Genome-wide risk prediction of common diseases across ancestries in one million people

Mars, Nina; Kerminen, Sini; Feng, Yen‐Chen Anne; Kanai, Masahiro; Läll, Kristi; Thomas, Laurent F.; Skogholt, Anne Heidi; Parolo, Pietro Della Briotta; Neale, Benjamin M.; Smoller, Jordan W.; Gabrielsen, Maiken Elvestad; Hveem, Kristian; Mägi, Reedik; Matsuda, Koichi; Okada, Yukinori; Pirinen, Matti; Palotie, Aarno; Ganna, Andrea; Martin, Alicia R.; Ripatti, Samuli

doi:10.1016/j.xgen.2022.100118

Cited by 57 publications

(53 citation statements)

References 77 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Our study is limited to individuals with European ancestry. Given recent discoveries about imperfect transferability of PRS between populations (Martin et al 2019; Mars et al 2022), training of an optimal CHDBioPRS for non-European ancestries would require appropriate training data from other ancestries. This is also important since the potential to use genetic scores to identify high risk individuals from birth could exacerbate the health differences between individuals with European ancestry and the others until a broader inclusion of underserved ethnicities in research become reality, particularly in multi-ethnic countries such as the UK and the US (Sirugo, Williams, and Tishkoff 2019).…”

Section: Discussionmentioning

confidence: 99%

Integration of biomarker polygenic risk score improves prediction of coronary heart disease in UK Biobank and FinnGen

Lin

Mars

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Background In addition to age and sex, also smoking history and levels of blood pressure, cholesterol, lipoproteins and inflammation are established biomarkers for coronary heart disease (CHD). As standard polygenic risk scores (PRS) have recently proven successful for CHD prediction, it remains of high interest to determine how a combined PRS of biomarkers (BioPRS) constructed from statistically relevant biomarkers can further improve genetic prediction of CHD. Methods We developed CHDBioPRS, which combines BioPRS with PRS of CHD, via regularized regression in UK Biobank (UKB) training data (n = 208,010). The resulting CHDBioPRS was tested on an independent UK Biobank subset (n = 25,765) and on the FinnGen study (n = 306,287). Results We observed a consistent pattern across all data sets where BioPRS was clearly predictive of CHD and improved standard PRS for CHD when the two were combined. In UKB test data, CHDPRS had a hazard ratio (HR) of 1.78 (95% confidence interval 1.67-1.91, area under the curve (AUC) 0.808) and CHDBioPRS had a HR of 1.88 (1.75-2.01, AUC 0.811) per one standard deviation of PRS. In FinnGen data, HR of CHDPRS was 1.57 (1.55-1.60, AUC 0.752) and HR of CHDBioPRS was 1.60 (1.58-1.62, AUC 0.755). We observed larger effects of CHDBioPRS in subsets of early onset cases with HR of 2.07 (1.85-2.32, AUC 0.790) in UKB test data and of 2.10 (2.04-2.16, AUC 0.791) in FinnGen. Results were similar when stratified by sex. Conclusions Integration of biomarker based BioPRS improved on the standard PRS for CHD and the gain was largest with early onset CHD cases. These findings highlight the benefit of enriching polygenic risk prediction of CHD with the genetics of associated biomarkers.

show abstract

Section: Discussionmentioning

confidence: 99%

Integration of biomarker polygenic risk score improves prediction of coronary heart disease in UK Biobank and FinnGen

Lin

Mars

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Moreover, only few studies have used genome-wide PRSs, although these contemporary PRSs containing a large number of variants have demonstrated improved performance beyond PRSs with less variants due to high polygenicity in common diseases. 13,28-30 Here we study the interplay of first and second-degree FH, parental causes of death, and genome-wide PRSs for 24 diseases using FinnGen (N=306,418), showing that FH and PRSs are largely independent and provide complementary information in risk assessment.…”

Section: Introductionmentioning

confidence: 99%

Systematic comparison of family history and polygenic risk across 24 common diseases

Mars

Lindbohm

Parolo

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Family history is the standard indirect measure of inherited susceptibility in clinical care, while polygenic risk scores (PRS) have more recently demonstrated potential for more directly capturing genetic risk in many diseases. No studies have systematically compared how these overlap and complement each other across common diseases. Within FinnGen (N=306,418), we leverage family relationships, up to 50 years of nationwide registries, and genome-wide genotyping to examine the interplay of family history and genome-wide PRSs. We explore the dynamic for three types of family history across 24 common diseases: first- and second-degree family history, and parental causes of death. Covering a large proportion of the burden of non-communicable diseases in adults, we show that family history and PRS are independent and not interchangeable measures, but instead provide complementary information of inherited disease susceptibility. The PRSs explained on average 10% of the effect of first-degree family history, and first-degree family history 3% of PRSs, and PRS effects were independent of both early- and late-onset family history. The PRS stratified the risk similarly in individuals with and without family history. In most diseases, including coronary artery disease, glaucoma, and type 2 diabetes, a positive family history with a high PRS was associated with a considerably elevated risk, whereas a low PRS compensated completely for the risk implied by positive family history. This study provides a catalogue of risk estimates for both family history of disease and PRSs, and highlights opportunities for a more comprehensive way of assessing inherited disease risk across common diseases.

show abstract

“…Assessing polygenic risk scores (PRS) transferability across ancestries is crucial if genetic risk stratification at the clinical level is to be implemented in populations with ancestry profiles that differ from those assessed in genome-wide association studies (GWAS) 1 . PRS derived by using individuals from a given genetic ancestry yield reduced predictions when tested in individuals of other ancestries, including those with admixed genomes, due to a complex interaction between differences in linkage disequilibrium (LD) and allele frequencies (AF) among populations, as well as population specific gene-environment and epistatic effects - all of which contribute to each disease or trait’s genetic architecture.…”

Section: Introductionmentioning

confidence: 99%

“…Breast cancer (BC) is the worldwide leading cause of death in women and a complex disorder of multifactorial inheritance. While up to 10% of BC cases are attributed to large effect pathogenic variants segregating in families, most affected individuals are not carriers, suggesting that sporadic occurrence is influenced by a polygenic combination of small to moderate effect variants 1 with familial aggregation and heritability reaching 55% 9 .…”

Section: Introductionmentioning

confidence: 99%

“…The potential use of PRS for improvements in disease prevention and management has been a prolific area of scientific investigation in the past decade, enabled by the availability of large datasets such as the UK BioBank (UKBB). While an essential resource, the UKBB and other major Biobanks are heavily biased in their ancestry representation 1,3,8 . Thus, much of the aforementioned assessments of transferability of PRS across ancestries rely on smaller cohorts that contain a majority of underrepresented ancestry groups such as Latino/Latinx/Hispanic, African/Black/African-American, Native American, and East Asian.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Assessing the risk stratification of breast cancer polygenic risk scores in two Brazilian samples

Barreiro

Almeida

Gomes

et al. 2022

Preprint

View full text Add to dashboard Cite

Polygenic risk scores (PRS) for breast cancer (BC) have a clear clinical utility in risk prediction. PRS transferability across populations and ancestry groups is hampered by population-specific factors, ultimately leading to differences in variant effects, such as linkage disequilibrium (LD) and differences in variant frequency (AF-diff). Thus, locally-sourced population-based phenotypic and genomic datasets are essential to assess the validity of PRS derived from signals detected across populations. Here, assess the transferability of a BC PRS composed of 313 risk variants (313-PRS) in two Brazilian tri-hybrid admixed ancestries (European, African and Native American) whole-genome sequenced cohorts. We computed 313-PRS in both cohorts (n=753 and n=853) versus the UK Biobank (UKBB, n=264,307) as reference. We show that although the Brazilian cohorts have a high European (EA) component, with AF-diff and to a lesser extent LD patterns similar to those found in EA populations, the 313-PRS distribution is inflated when compared to that of the UKBB, leading to potential overestimation of PRS-based risk if EA is taken as a standard. Interestingly, we find that case-controls lead to equivalent predictive power when compared to UKBB-EA samples with AUROC values of 0.66-0.62 compared to 0.63 for UKBB.

show abstract

Genome-wide risk prediction of common diseases across ancestries in one million people

Cited by 57 publications

References 77 publications

Integration of biomarker polygenic risk score improves prediction of coronary heart disease in UK Biobank and FinnGen

Integration of biomarker polygenic risk score improves prediction of coronary heart disease in UK Biobank and FinnGen

Systematic comparison of family history and polygenic risk across 24 common diseases

Assessing the risk stratification of breast cancer polygenic risk scores in two Brazilian samples

Contact Info

Product

Resources

About