One of the biggest challenges in drug discovery for neurodegenerative diseases is the large heterogeneity that characterises the patient population. Motor neurone disease (MND) is no exception. Patients can present the disease in different ways, with symptoms starting in different parts of the body, different ages, the disease progressing at different rates. This clinical heterogeneity is further complicated by individual genetic characteristics. All this variability is likely to underline the involvement of different mechanisms involved in neuronal death.
Decades of research have produced large amounts of clinical and experimental data that we can use to classify MND patients better, based on their clinical similarities and differences, and to identify mechanisms involved in disease. Such effort requires the use of cutting-edge technologies such as artificial intelligence, which have the ability to assimilate large amounts of information coming from different sources, including scientific literature, clinical records and chemistry data to match disease mechanism and appropriate drugs.
In fact, we believe that speeding up the process of drug discovery is of paramount importance for ALS patients and their families. For this reason, in our lab we work in a collaboration with BenevolentAI, a biotech company that uses artificial intelligence with an aim to improve the way new ALS therapeutics are discovered and patients are classified. Thanks to their access to cutting-edge AI and machine learning technologies, BenevolentAI are able to search the entire breadth and depth of all scientific information ever published and look for new, unexplored links between pathological processes in ALS and potential new drugs. Drugs that are shortlisted as best candidates are then tested by our lab in a high-throughput co-culture screening using patient-derived cells to determine which therapeutic agent has the best effect on survival of motor neurons. The top-performing drug is then characterised by us in terms of its effect on patient cells, to understand fully what the compound does in ALS before proceeding to safety testing and eventually clinical trials.
All in all, harnessing AI technologies could not only help us build a comprehensive picture of ALS, but also explore avenues to new treatments.
If you would like to find out more about the ways AI can be used to advance diagnostic and therapy measures in neurodegeneration, have a look at our recent article.