Future of medicine

“Every decision about your health is based on population-based measurements. But everybody’s baseline is different.”

Precision personalised medicine

In the future, a visit to the doctor could involve taking a look at medical records that include wearable data, collected over many years and combined with a detailed personal omics profile. ‘It’s not meant to replace your physician, it’s meant to work with your physician,’ says Snyder. Automatic data analysis conducted by algorithms trained on reams of anonymised patient data could also help to spot patterns and outlier patients before they even realise they are ill, referring them to their doctor for further tests and possible treatment. A version of that future is already available – if you’re wealthy enough. In 2015 Snyder co-founded a company called Q Bio, which promises to provide people with a comprehensive picture of their health, based on big data analytics. At $3,500 a visit, it’s not cheap, but Snyder is confident the cost will come down considerably over time. And as costs come down, a version of Snyder’s omics profiling could become the new standard for healthcare providers, transforming doctors’ surgeries from places we visit when we get sick to places we visit to keep us healthy

We’re collecting all kinds of sophisticated information, pulling it back, getting it processed into a fashion that both the physician and the consumer can understand.’ Get that part wrong and the data risks becoming overwhelming or meaningless, even if it is accurate.

Covid-19 has changed very little as it has spread across the world. But, in infecting people of all ages and races, the disease has laid bare the variation between humans, and shown the flaws in a one-size-fits-all approach to healthcare. ‘It’s the world’s largest and saddest clinical trial. But the one thing we can do with it is try to do our absolute best to improve for the future,’ says Regev. ‘It’s played out completely differently in different individuals. And we have to know why, if we want to control it.’

Living drug

Reprogramming immune cells into ferocious cancer-killers. This is the very definition of precision, personalised medicine: a patient’s own cells, re-engineered to kill the cancer that has taken over their bodies, potentially giving them a second chance at life when almost all hope is lost.

AI

Faced with Covid-19, the researchers at BenevolentAI had to recalibrate the system to look for a potential treatment that already existed – a so-called ‘off-label use’ for a drug that was approved for use in humans. ‘We can find new connections,’ says Alix Lacoste, vice-president of data science at BenevolentAI. ‘It would take a human a really long time, because you have to sift through billions of interactions. But a machine is a lot faster.’ These interactions are chunks of information on a vast database known as a knowledge graph – essentially a collection of more than a billion relationships between genes, targets, diseases, proteins and drugs. To search this database, BenevolentAI’s researchers use a toolbox of specially trained AI assistants with the uncanny knack of finding needles in a veritable haystack of pharmaceutical data. The premise is simple: among the troves of academic papers and literature there must be scores of significant discoveries that have been forgotten about or overlooked. Put an AI to work on that data, however, and new discoveries will reveal themselves.