But single-gene inheritance is only part of the picture being uncovered by Professor Craig and his team, including inaugural Snow Fellow, Associate Professor Owen Siggs.

Their program has recruited thousands of patients with severe glaucoma to identify more common and subtle genetic changes that may contribute to the disease.

“By doing that, we’ve come up with around 2,500 small genetic contributions to glaucoma that additively can make a very solid prediction for an individual,” he says. 

Associate Professor Siggs adds “In some circumstances, these predictions are as good as genetic tests we already use in the clinic today.”

That research was published last year in Nature Genetics and allows doctors to ascribe a “polygenic risk score” (or PRS) to patients.

“In doing so, we can make a really strong prediction. People in the highest 10% of risk are 15 times more likely to get glaucoma than people in the lowest 10% of risk,” he explains. 

That means a blood or saliva test could give a good indication of whether an individual is in a very high or low risk group.

While that doesn’t address the middle groups, the research has thrown up some unexpected findings that may make it more universally adaptable, such as whether a high-risk patient will need surgery, and at what age they get the disease. It also predicts how many family members are likely to get glaucoma.  

“So now we’re designing a test that can be performed as a routine screening, similar to the over-50s bowel cancer screening.

“If we could access a sample from everyone at the age of 50, we would identify those in the very high-risk groups, and advise they turn up for a regular eye check-up for glaucoma.”

The team is also running a longitudinal research project monitoring around 2,000 people every six months. That has been going on for more than five years and has revealed a lot about who loses vision and how rapidly. 

“It turns out that our PRS test is also quite good at predicting the rate of progression in early disease and who ends up losing vision, requiring treatment.

“We think that it might have value amongst people who have a borderline eye pressure test, for example, as then an optometrist could get those people back for further observation.”

But Professor Craig warns that eye pressure tests are not always a completely reliable indicator – around 30% of people with glaucoma have normal pressure. 

While regular imaging of the optic nerve as well 

as laser scanning can get around that and is highly effective at picking up early-stage glaucoma, it also picks up people who don't have glaucoma and are unlikely to contract it. 

If all those with some borderline features were referred to a specialist, it would threaten to swamp the system.

“It's what we call the false positive conundrum,” says Professor Craig. “By doing really sophisticated testing, you can end up with a massive number of patients to deal with.”

The answer may lie in precision medicine made possible by artificial intelligence.

“We are working on the use of AI to refine our predictions and to work out who actually has a problem. In what we call the PROGRESSA study, the data includes laser photographs, pressure ratings and a whole lot of other measures about the rigidity of the eyeball, performance on field tests and the genetics. 

“We’re currently feeding all of that information into AI algorithms. That's quite exciting because it takes it towards precision medicine where we can apply treatments, screening and monitoring at the individual level rather than just one size fits all.”