Tartu scientists showed that people’s descent plays important role in genetic risk estimation

In a recently published article, Tartu scientists showed that unless an individual’s genetic descent is taken into account, his or her DNA-based estimation of disease risk may provide incorrect results.

In the case of many complex diseases, for example the coronary heart disease and the type 2 diabetes, the disease risk is not affected by one or two, but thousands or even tens of thousands of gene markers available in the human hereditary substance, DNA. Each one of them has a relatively small effect if taken separately but the aggregate impact is considerable.

This is why different mathematical models have been developed in the world and also in Estonia, attempting to take into account all these gene markers and sum up the combinations that increase the specific person’s risk and, based on the obtained result – the so-called risk score – determine the person’s disease risk as low, medium or high.

“In the article we showed that the thresholds required for risk estimation are not universal, but are significantly dependent on the genetic descent of the person. A risk score that means a high probability of the expression of a disease for an Estonian, may be a low one for an African,” one of the authors of the article Sulev Reisberg explained the finding. Reisberg added that therefore it is extremely important to adjust the score according to the person’s descent before giving the risk estimation. “This may prove really complicated if the individual is not a typical representative of the population genetically, and if their roots are mixed. Often the person may not even know his or her exact descent.”

Authors of the article do not say how precisely to adjust the risk score – this is a topic for further research. “It was important to first of all estimate the possible effect. When we know the problem, we can start looking for a solution – either by eliminating a part of the people from risk estimations to avoid giving them incorrect feedback, or by adjusting the score to match their genetic background,” said leader of the research group Professor Jaak Vilo.

Authors of the article are scientists from the University of Tartu Institute of Computer Science, Estonian Genome Centre, Software Technology and Applications Competence Center and Quretec.

The article was published in PLOS ONE journal and the original text is available here: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0179238

Additional information: Sulev Reisberg, PhD student, +372 524 8123, sulev.reisberg@ut.ee