Home   News   Article

Subscribe Now

University of Cambridge researchers to use machine learning to find type 1 diabetes treatments




Researchers at the University of Cambridge have been given a share of £2million in funding from the Type 1 Diabetes Grand Challenge to identify and test new drug candidates for the condition.

Professor Eoin McKinney and his team will use genetic insights and machine learning technology.

Monitoring blood glucose levels in a diabetes patient
Monitoring blood glucose levels in a diabetes patient

Type 1 diabetes is an autoimmune condition affecting 4,320 people in Cambridgeshire in which the immune system mistakenly attacks and destroys insulin-producing beta cells in the pancreas.

We can detect signs of the immune system’s attack in the blood long before symptoms of type 1 diabetes start. Known as autoantibodies, these signs are present in the early stages of the development of the condition and screening studies - like ELSA for children and T1DRA for adults - can test for them.

Those testing positive can be offered immunotherapy treatment as part of a clinical trial.

However, there are currently no licensed immunotherapy treatments for the condition and a raft of them will be needed to everyone with type 1 diabetes autoantibodies.

Prof McKinney’s team has already created detailed maps of immune cell changes in people at high risk of type 1.

They have identified specific patterns, known as signatures, seen only in those who develop the condition and now aim to better understand the abnormal immune signals that lead to it and find drugs to counter them, preventing the condition from progressing.

They will investigate whether any of these type 1 signatures match drugs already approved for other conditions, then test if they help prevent the immune system attack in cells in the lab and in mice.

This approach, using drugs already known to be safe for other uses, will speed up the drug discovery process and identify treatments that can protect beta cells and insulin production.

Prof McKinney said: “By collaborating with international groups, we have generated detailed maps of immune cell changes occurring from the earliest stages of T1D through to diagnosis.

“Now, we plan to use that information to find new treatments that might prevent progression, by matching changes seen before type 1 diabetes to those produced by commonly used medicines. By finding a match, we hope to identify which drugs might be most easily used to reverse or stop disease progression.

“By selecting candidate treatments rationally based on a match with type 1 data, we will stand the best possible chance of finding a safe and effective approach to stop the condition with real impact for patients everywhere.”

Rachel Connor, director of research partnerships at Breakthrough T1D, said: “Prof Eoin McKinney’s project looks to identify potential new drugs to treat type 1 diabetes by harnessing the power of machine learning. Applying this advanced technology to type 1 diabetes is incredibly exciting and could lead us to new treatments faster than ever before. Slowing and ultimately stopping the immune attack that drives the development of type 1 is critical to delivering cures to people living with this condition.”

Dr Elizabeth Robertson, director of research at Diabetes UK, said: “By leveraging genetic insights and repurposing existing drugs, Prof McKinney’s team is pushing the boundaries of what’s possible in preventing the progression of type 1 diabetes. This research holds huge promise for changing the lives of those living with or at risk of this condition, accelerating us towards our ultimate goal of finding a cure.”

The Type 1 Diabetes Grand Challenge is a partnership between the Steve Morgan Foundation, Diabetes UK and Breakthrough T1D (formerly JDRF). The funding comes from a £50million investment from the Steve Morgan Foundation focused on type 1 diabetes.

Imperial College London, Oxford and Exeter are also sharing in the £2m funding.



This site uses cookies. By continuing to browse the site you are agreeing to our use of cookies - Learn More