CAMBRIDGE, U.K.—Researchers from the Wellcome Sanger Institute, the Broad Institute of MIT and Harvard, and colleagues from 101 research institutions world-wide have studied hundreds of thousands of participants and identified over 7,000 regions of the human genome that control blood cell characteristics, like the numbers of red and white cells. Two large-scale genetic studies have now identified the bulk of genetic variations that influence medically-important characteristics of our blood cells.
A paper published today in Cell now shows how a person’s genetic make-up contributes to them developing blood diseases. This knowledge brings us one step closer to using genetic scoring in the clinic to predict personal risk of developing blood disorders.
“In this study, we have been able to show how a person’s genetic predisposition to certain blood-related measurements, as indicated by their polygenic score, can predispose them to blood disease,” said Dr. Dragana Vuckovic, Wellcome Sanger Institute, NIHR Blood and Transfusion Research Unit in Donor Health and Genomics at the University of Cambridge and a first author of the study. “If a person is more genetically predisposed to low hemoglobin, for example, then they are more likely to develop anemia.”
Blood disorders like anemia, hemophilia and blood cancers are a significant global health burden. Many of these disorders can be viewed as extremes of normal biological states — as in anemia, where having too few red blood cells results in inadequate oxygen supply to the body. These extremes can occur as a result of small variations in our DNA, some of which increase our risk of developing a disease.
By comparing the DNA sequences of large numbers of individuals, it’s possible to investigate how genetic variations translate into physical characteristics or traits. This includes the chance of developing common diseases like asthma, heart disease and hemophilia.
In these studies, anonymized genomic and healthcare data from the UK Biobank and other studies from the Blood Cell Consortium were analyzed. Study participants included people of European, East Asian and African American ancestry. The authors discovered 7,193 distinct genetic regions associated with 29 blood cell measurements, which represents the largest set of correlated genetic regions identified to date.
The researchers assessed the potential for predicting blood cell traits based on polygenic scores. They found that polygenic scores could predict predisposition to complex diseases, including blood disorders.
“The construction of polygenic scores requires the analysis of large amounts of data. Our study shows that the performance of polygenic scores for predicting blood cell traits is improved by careful selection of a smaller set of genetic associations determined by deeper statistical analyses of the available data. This finding disrupts a common assumption that including a greater number of genetic associations will result in a better predictive polygenic score,” added Parsa Akbari, University of Cambridge and a first author of the study.
“[T]he emerging picture of underlying network connectivity regulating blood traits harbors potential for discovering new pathogenic genes and drug targets,” the article states. “As an example, we identified a subset of 11 closely coexpressed genes (including three known platelet genes (GP9, ITGA2B, and GP1BB) that is coregulated by the same trans-acting eQTL in the ARHGEF3 gene. The use of a large dataset with concurrent genetic and gene expression data in different cell states will be necessary for further quantitative validation of this model (Liu et al., 2019).”
“Polygenic variation has a substantial contribution to variation in complex quantitative traits and disease risk, sometimes yielding effects comparable to those of rare pathogenic variants. Using only the sentinel signals from our discovery GWAS [genome-wide association studies], we built PGSs [polygenic scores] explaining up to 28% of phenotypic variance,” notes the article. “We explored the polygenic effects jointly with pathogenic variants and as phenotype modulators in patients with rare blood disorders.
“While we found that 16 known monogenic variants were each associated with quantitative blood traits, 52 participants homozygous for five rare recessive pathogenic variants appeared to be healthy with normal blood count and indices. This suggests that the penetrance of pathogenic variants may be overestimated in many instances, as was recently shown (Oetjens et al., 2019). Differences in PGS could not explain the reduced penetrance, but our analysis may be limited by the diseases we had adequate statistical power to assess.”
“Conversely, we observed strong allele dosage-dependent effect sizes for two heterozygous variants (previously reported as recessive), that could lead to disease especially if coinherited with an adverse PGS. Finally, we observed a significant association between phenotype-relevant PGSs and rare blood disorders for thrombocytopenia, secondary polycythemia, anemia, and aplastic anemia, regardless of the presence or absence of known rare variants in patients,” the article concludes. “This highlights a substantial polygenic modulating effect on presumably monogenic disorders and lays the groundwork for future studies aiming to define the impact of genetic background on the variable penetrance and expressivity in blood disorders.”
As well as variations in DNA, environmental and other factors are involved in complex diseases like asthma or hemophilia. Analyses conducted in this study improved the performance of polygenic scores, increasing their potential as a powerful tool to help predict personal risk.
“This study indicates that polygenic scores could be used routinely in personalized medicine in future, following further research. The DNA of every human being contains millions of variations that make us unique and which influence what is ‘normal’ for each of us,” mentioned Professor Nicole Soranzo, lead author of the study from the Wellcome Sanger Institute and University of Cambridge. “Genetics now helps us to benchmark what is ‘normal’ from birth, and allows us for the first time to monitor for deviations from this baseline that might indicate an increased risk of disease during our lifetime.”