Scientists learn to predict diseases by the human genetic code

A combined group of American mathematicians and biologists has developed algorithms for processing polygenic models. To the general public, they are known as "genetically determined diseases" or ailments provoked by the activity of specific genes. Only now, in practice, there is an unimaginably huge number of combinations of these genes, and only now scientists have roughly understood how to take into account the influence of all of them when making a diagnosis.

One gene, no matter what function it is responsible for, rarely becomes the cause of the disease, almost all diseases are inherently polygenic - provoked by the activity of groups or subsets of genes, which are often not connected at all. American scientists have developed a model that can accommodate a very large number of different combinations of genes. And they applied an evaluation algorithm based on the "big data" analysis methods to it, having received a frighteningly accurate tool.

As an example, the study of the probabilities of coronary artery disease, which depend on 6, 6 million positions in the genome. Based on data on patients from the British medical database Biobank, the authors of the methodology concluded that for 8% of people the likelihood of such a disease is three times higher than for all others. The lower risk level according to the new system leaves you with only 1 chance in 100 to get sick, but the higher index guarantees the disease in 11% of cases.

It seems that with such a spread, the accuracy of the forecast is out of the question, but look at the situation from the other side. Data can be obtained on the basis of DNA analysis for any person, starting from the first minutes of life. And make a plan - in what conditions it is better for him to live, what to eat and drink, what medications to take in order to never get sick with something specific. For a businessman, sports trainer, astronaut recruiter, the 10% difference is already an argument to make a choice between candidates not to spend decades and efforts on training a less promising person. And how happy the insurance companies will be with their dynamic tariffs ... that's why the technology is not yet thought to be implemented. Keep it within the laboratory.