Population genetics explores genetic differences both within and between populations and individuals over time. This primarily encompasses the quantitative study of genotypes and phenotypes within the context of historical demography. Improved next generation sequencing technologies, increased scalability in genetic data processing, coupled with impressive global data sharing initiatives provide unprecedented research opportunities, allowing us to understand the links between ancestry, genetics and health outcomes.
Our center’s overall goal is to characterize human genetic ancestry in order to inform better models for clinical medicine.
Worldwide frequency of the Steel Syndrome variant (Belbin et al., eLife, 2017)
We achieve this using advanced statistical and probabilistic modeling and inference to determine genetic associations, subpopulations and relatedness. The genetic diversity observed in modern metropolitan communities, such as New York City, provides an exciting opportunity to improve the resolution of human populations at a genetic level. These are crucial steps in improving functional gene discovery in global populations. This in turn will further our understanding the genetic basis of disparities in disease prevalence and burden across populations and within the the healthcare system.
Exploratory paradigms in health research science
Exploratory genetic studies look for statistical relationships between genetic differences in people and the various traits that they have. One of the most common approaches to accomplish this is a genome-wide association study (GWAS): where we look for a genetic variant (or variants) that occurs more frequently in people with a particular disease than in people without the disease. We can also switch this around: focusing on genetic variants that are suspected to be clinically relevant, then investigating whether these variants influence any of the observable characteristics, called phenotypes, that are captured in electronic health records. This method is known as a phenome-wide association study (PheWAS). Associations from these sorts of exploratory tests can then be evaluated in a more focused study, and be used to improve clinical care.
Identity-by-Descent for Fine-scale Ancestry Inference
We are interested in exploring demographic patterns using segments of the genome inherited identically between individuals from a recent, shared common ancestor. These shared segments of the genome are known as segments inherited Identical-By-Descent (i.e IBD), and can be used to re-construct population-scale pedigrees of genetic relatedness between many individuals. Exploring patterns within these pedigrees can allow us to identify groups of people that are statistically enriched for the sharing of many recent, common ancestors. This genealogical characteristic is strongly correlated with recent, shared fine-scale geographical or cultural origin. Recent shared ancestry is important from a genetic standpoint because many variants of clinical interest tend to be extremely rare globally, but are sometimes more common than expected at a local scale among people who share recent ancestry.
The characteristic of rare, clinically relevant variants segregating at high frequencies locally can be further exacerbated when populations have undergone demographic processes such as the rapid reduction in the number of individuals in a population, leading to reduced genetic diversity (a phenomenon known as a founder effect). By quantifying patterns of IBD sharing within groups, it is possible to make inferences about whether these processes have played a role in their demographic history, and in turn the likelihood that these populations may harbour disease-causing variants at a frequency that may inform public health policy (i.e. population screening).
Phenome-Wide Exploration of Health Disparities using Electronic Health Records
Population-level disparities in diseases and drug response are known to exist between groups for a variety of diseases. The reasons for health disparities are complex, with many potential underlying causes, including environmental, socio-economic and cultural factors, as well as genetics. Understanding health disparities between populations has important implications for health care and health policy. Historical approaches to medicine and healthcare have not considered the need for a personalized approach to account for such differences in etiology. One aspect of research within the centre seeks to identify patterns of enrichment of clinical phenotypes within the Electronic Health Records
(EHR) of BioMe participants at Mount Sinai. An example of this is “Ancestry PheWAS” analysis which identifies health disparities in an automated fashion by measuring statistical enrichment of medical billing (ICD-9) codes related to clinical phenotypes within fine scale ancestral communities. This community centric approach can find enrichment of disease phenotypes in a given ancestry. Following this further association testing based on ancestry haplotypes uncover if there is a genetic contribution to the observed differences in disease prevalence.
Biobank-scale real-time pharmacogenomics
Through collaborations with medical departments at the Mount Sinai Hospital, we’re investigating the genetic basis underlying differences in responses to drugs commonly administered during surgery, with the ultimate goal of providing patients with personalized recommendations based on their genetic profile.
As drugs interact with each other but are also impacted by patients medical history and clinical characteristics, we take into account every patient’s full electronic health record as well as their genetic profile. We combine all of this information through state-of-the-art statistical genetics methods, which allow us to highlight differences between patients, show how different groups of patients react strongly to certain drugs, and customize treatments for every patient.