Are some ethnicities being inadvertently left behind in the genomic revolution?

Simon Odekunle

Recent scientific and technological breakthroughs, such as the advent of CRISPR gene-editing techniques, have brought the concept of genetics-based medicine (GBM) closer to reality quicker than most predicted. Based on the information gained from the multitude of genomic research projects taking place all over the world – and set to take over the world at some point in the near future, in the form of personalised medicine – there is still one issue being overlooked by many scientists researching genetic diseases: sampling bias (Konkel, 2015). This is particularly true of ethnic diversity in genomic studies. While most researchers are aware of the existence of genetic variation between people of different ethnicities, little effort is being made to include a wider variety of ethnicities in genomic studies.

In order to understand how this bias arises, and why this is such a major issue, there are two major questions that must first be addressed:

  1. How are these genomic studies being carried out?
  2. How can sampling bias negatively impact the development of drugs and therapeutic methods?

Most GBM is based on the results of genome-wide associated studies (GWAS). GWAS are studies in which the genomes of large populations of people are analysed, in the hopes of identifying genetic differences in regions of the genome associated with genetic diseases – such as Alzheimer’s, asthma and type 2 diabetes – which could lead to the discovery of new therapeutic approaches for these diseases (Konkel, 2015). Many genetic diseases are associated with specific single nucleotide polymorphisms (SNPs) in important regions of DNA. These SNPs can render the proteins encoded by various genes dysfunctional in a way that may increase a person’s susceptibility to a particular disease. GWAS have helped revolutionise how researchers study and treat certain diseases through the identification of SNPs of particular importance, and has greatly advanced the way researchers view and treat diseases like Alzheimer’s and atherosclerosis (Boyer, 2011).

However, it is in the identification of these SNPs that the problem lies. These large sample populations are often either fully or mostly comprised of individuals of European descent, with very little representation for those of different ethnic origins (Bustamante et al., 2011).

e

Ethnic differences among participants in GWAS (based on data from Bustamante et al.)

This is an important issue, because SNPs do not arise solely due to mutations related to disease but also as a natural response to one’s environment. As an individual or a group of individuals live and adapt to certain conditions, more SNPs arise, distinguishing these individuals from those who have not lived under those conditions. This means that a person who lives in Asia, for example, and whose ancestors grew up in that same environment, will have a slightly different set of SNPs than someone who was born and raised in the Americas

SNPs can determine how researchers identify genetic diseases, and so possessing SNPs that have no influence on a person’s susceptibility to a disease – but rather, are neutral differences that arise merely as a result of environmental factors – can lead to some unforeseen consequences. These environmental SNPs can mask the locations of disease-related SNPs, making it more difficult to identify the same SNPs in individuals whose ethnicity’s SNP profile may not be so well documented. These variations can lead to numerous complications when a therapy designed using the SNP profile of one ethnicity, is used to treat an individual of another ethnicity.

For example, in dozens of studies done on type 2 diabetes in European populations, 19 SNPs were identified to be strongly linked to the disease (Bustamante et al., 2011). However, a study that took a sample size of 6,000 individuals of various ethnicities – such as Latino, European American, African American, Native Hawaiians, and Chinese- and Japanese American – found that only 13 of these SNPs continued to be strongly associated with the disease, and 5 of these SNPs had different effects on an individual’s health depending on their ethnic background (Bustamante et al., 2011). This variation could mean that any drug made to treat type 2 diabetes based on those 6 non-uniform SNPs may not have the desired effect on people whose genes do not match the archetypal genome used to design the drug. This is not a rare occurrence. As researchers perform broader-spectrum GWAS, it is becoming more apparent that it may not be safe to presume that disease-related SNPs identified in one ethnicity have the same significance in other ethnicities.. This is particularly true of atherosclerosis (a hardening and narrowing of the arteries) whose causal SNPs seem to vary between ethnicities (Boyer, 2011).

There is no easy remedy to this issue. Although many organisations are setting guidelines for GWAS and incentivising the researchers conducting them to be more inclusive of individuals from all ethnic backgrounds, more needs to be done. An effort needs to be made to go after samples from individuals of varying ethnicities, and to replicate monoethnic GWAS in other regions of the world. Yes, this may be costly and oftentimes inconvenient (and little glory comes from being the second to do something), but that does not mean that significant contributions to science cannot be made. Other possible advantages to resetting this imbalance in GWAS include: promoting global health equality, prompting members of local diasporas to participate in GWAS, developing novel drugs and therapies to accommodate these different SNP patterns between ethnicities. There are many ways to ensure that all people from all ethnicities are included in the genomic revolution. These steps must be taken if the true goal of biomedical researchers is to provide the world with effective drugs and therapies for as many people and as many diseases as possible.

References:

  1. Boyer, T. (2011) Heart Disease Studies: Is There a Racial Bias Between Blacks and Whites? [online]
  2. Bustamante, C., Burchard, E. and De La Vega, F. (2011) Genomics for the world. Nature. 475(7355), 163–65
  3. Haga, S.B. (2009) Impact of limited population diversity of genome-wide association studies. Genetics in Medicine. 12, 81-84.
  4. Konkel, L. (2015) Racial and Ethnic Disparities in Research Studies: The Challenge of Creating More Diverse Cohorts [online]
Advertisements