From polygenomics to omnigenomics

Biologists have questioned the benefits of large-scale genetic research

Daria Spasskaya, N+1

Geneticists from Stanford University in California have suggested that the value of large-scale genetic research aimed at finding associations between inherited traits or diseases and single-nucleotide polymorphisms may be exaggerated. Article by Boyle et al. An Expanded View of Complex Traits: From Polygenic to Omnigenic is published in the journal Cell, and the authors' comments can be found in a note in Nature (New concerns raised over the value of genome-wide disease studies).

Genome—Wide Association Studies (GWAS - Genome-Wide Association Studies) — this is a field of genetic research that establishes a connection between any traits and genetic markers. The latter are most often single—nucleotide polymorphisms (SNP - Single-Nucleotide Polymorphisms), that is, variants in the DNA sequence that occur quite often in the population (more than one percent of people). The studies are based on comparing the genomes of a group of people with a given trait or suffering from a disease with a control group. Based on statistical processing of the frequency of occurrence of certain variants in two groups, it is concluded that variant X (for example, nucleotide A in the gene sequence) is associated with trait Y (for example, obesity).

Such studies have become widespread with the increasing availability of DNA sequence analysis and have gained momentum significantly over the past ten years. If the first articles on the topic of genetic associations analyzed samples of several hundred people, then in 2017 a decent sample size is 200 thousand or more. For example, a million people participate in the study of type 2 diabetes conducted by the University of Oxford. Such volumes of data are necessary to make conclusions more reliable and take into account a larger number of genetic variants.

Thanks to large-scale research, a large amount of valuable data has been obtained. Different variants of proteins involved in the metabolism of drugs in the liver have suggested to doctors that patients may respond differently to drugs, and in some cases the dose should be varied individually. Other studies have made it possible to clarify the mechanism of disease development, for example, to take into account the contribution of the central nervous system to the development of hereditary obesity.

However, despite the obvious benefits of genetic associations, most often their results are difficult to interpret. Most of the detected variants have a very weak, although statistically reliable relationship with a given trait. The existence of a link does not yet imply a mechanism by which gene X affects the trait Y. Often X refers to factors that at first glance do not relate to Y. at all. Roughly speaking, large-scale studies can lead to the conclusion that "everything affects everything."

Geneticists from Stanford offered a critical look at the genome-wide analysis of associations and reanalyzed some published studies on this topic. For example, taking data from a 2014 paper on the genetics of such a complex trait as height, the authors found that 62 percent of all human polymorphisms are somehow associated with growth, although a reliable connection can be established only for three percent. Each of the polymorphisms conditionally contributes to the growth of one and a half millimeters.

Associations of polymorphisms with growth. On the X – axis is the value,
reflecting the reliability of the relationship of the variant with the trait
(the smaller it is, the better). Drawings from an article in Cell.

Geneticists also studied data for schizophrenia and two autoimmune hereditary diseases — rheumatoid arthritis and Crohn's disease. For all three diseases, there are genetic associations with changes in gene expression in specific tissues (in neurons for schizophrenia and immune organs for autoimmune diseases). However, scientists have found that polymorphisms located simply in active regions of the genome, rather than in specific tissues, make a greater contribution to the genetics of these disorders. 

Graphs reflecting the contribution of polymorphisms affecting gene expression 
in individual tissues or in general in actively read parts of the genome.

Thus, the contribution of a particular genetic variant is most often indirect. For example, polymorphism can affect the work of a regulatory complex that controls hundreds of genes, and some of them may already be directly related to a disease or other trait. Based on this, the authors of the article propose to recognize that in reality everything affects everything, and move from the concept of "multigenic" traits to an "all-gendered" concept. In order to interpret the data already collected, it is necessary to focus on the analysis of intracellular regulatory networks, and not to conduct new, even more extensive studies. This is the only way to understand why variant X is associated with trait Y, and to draw practical conclusions, according to geneticists.

The authors remind that in the case of multifactorial traits, the contribution of genetic markers cannot be considered separately. They have value only in the form of the sum of many small changes. A search is underway for genome-wide associations even with such complex traits as reproductive behavior and intelligence.

Portal "Eternal youth" http://vechnayamolodost.ru  22.06.2017