��ҹ����1000����

An analysis of the genetic data of millions of people can help predict their risk of developing common diseases and could help the UK’s National ��ҹ����1000���� Service save lives and money, a firm spun out of the University of Oxford has claimed.

Genomics plc says it has produced polygenic risk scores for heart disease, breast cancer and 14 other diseases by examining more than 3 million people’s genetic data. Half a million genomes came from the UK Biobank, while the rest were from more than 200 other institutions.

Such scores could help the NHS take preventative action and target scarce resources better, says co-founder Peter Donnelly. For example, women whose genes indicate they have a higher risk of breast cancer could be screened earlier than their 50th birthday, when checks usually start.

��ҹ����1000���� secretary Matthew Hancock said today that polygenic testing , and he is due to give a speech later on increasing the use of such techniques in the UK. “We must get predictive testing into the NHS as soon as we possibly can,” Hancock will say. But can polygenic scores really help the health service?

Big data testing

Genetic testing is already used in the NHS to look for rare diseases caused by a single gene, such as the brain disorder Huntington’s disease.

Polygenic testing aims to predict the risk of more common diseases, such as coronary heart disease, which can be made more likely by hundreds of different genes.

Making such predictions used to be impossible, because each of the genes has a small influence on the risk of developing a disease. But analysing huge genetic databases makes it feasible to tease out these effects.

“This is a totally different use of genetics in healthcare,” says Donnelly. “I’m convinced this is where genetics will have its biggest impact.”

Anneke Lucassen at the University of Southampton, UK, is a cautious enthusiast for polygenic risk scores, which she believes could help identify high risk groups and play a role in targeting screening for diseases.

But she says it remains unclear what people who receive a polygenic risk score should do with the information. “We tend to assume that picking up something early makes it more treatable, but sometimes there is little evidence for that assumption and the downside of overdiagnosis is oft ignored.”

Other researchers are also sceptical about the technique’s accuracy. “For most diseases, the best polygenic risk scores do not perform well enough to be of practical use,” says Doug Speed at Aarhus University in Denmark. That view is shared by David Curtis at University College London. “The utility of the scores has been exaggerated,” he says.

That leads to another problem: there is no single, definitive polygenic risk score for any given disease. The scores that Genomics plc arrives at, for instance, may differ from those of another company, meaning the NHS would have to choose which one to use.

Polygenic risk scores are currently also less accurate for anyone whose ancestry is non-white European because they are underrepresented in the genetic databases the scores are derived from. “I think that’s a deal-breaker,” says Curtis.

Donnelly notes concerns that the scores don’t work well for ethnic minorities, but says the answer is to gather more data from those populations and to use new methods of analysis. He says the company will take advantage of genetic studies under way in different ancestry groups to produce tools that are equally effective for everyone.