Comment: How AI could accelerate breakthroughs in most common form of motor neurone disease
Artificial intelligence (AI) is increasingly being used to support medical research and drug discovery, but its effectiveness depends heavily on the quality, scale and diversity of the data available to researchers.
In this comment piece, Amalia Papanikolaou, Data Programme Lead at Challenge Works, discusses why unlocking and connecting ALS patient data is critical to accelerating research into amyotrophic lateral sclerosis, the most common form of motor neurone disease.
AI is rapidly reshaping how we work, think and solve problems. Across every sector, conversations around its potential are evolving at pace, and healthcare is no exception.
Among its many capabilities, AI is enabling scientists to analyse vast datasets to uncover biological patterns, transforming how diseases are understood, diagnosed and treated. Once largely experimental, AI is fast becoming a cornerstone of clinical decision-making and medical innovation, and its role in drug discovery is increasingly well evidenced.
However, when it comes to drug development, AI systems are only as powerful as the data they learn from. This is particularly true when developing treatments for complex diseases such as amyotrophic lateral sclerosis (ALS), the most common form of motor neurone disease (MND).
The complexity of ALS has long hindered progress towards meaningful treatments. However, years of dedication from global charities and research institutions have helped raise awareness and generate a wealth of ALS patient data that simply did not exist before. This progress is hugely significant, as the more data researchers can access, the greater the chance of developing a viable treatment.
But scale alone is not enough. For AI to deliver breakthroughs that benefit everyone, the data it learns from must also reflect the diversity of the populations it is intended to serve.
Historically, many communities have been underrepresented in medical research and clinical datasets, creating blind spots that risk reinforcing existing health inequalities. If we want AI to unlock more effective and equitable treatments, we must ensure that patient data includes people from different ethnic, geographic and socioeconomic backgrounds.
The challenge now is to continue pushing for access to as much high-quality data as possible. Globally, there must be a greater commitment to improving the scale, quality, diversity and connectedness of patient data. Only then will AI be able to meet its full potential.
The current outlook on ALS
Despite these challenges, there has been meaningful progress for people living with ALS in recent years.
The approval of Tofersen for the small proportion of patients with the SOD-1 gene mutation marks a significant breakthrough, while emerging research, from AI-designed molecules targeting disease-driving proteins to experimental imaging agents that could improve diagnosis, signals growing momentum across both treatment and detection.
But ALS remains incurable, and those diagnosed are given three-to-five years to live on average. Crucially, ALS is not a uniform disease. Its progression differs significantly between patients, with symptoms, rate of decline and survival varying widely.
That unpredictability is one of the greatest barriers to developing effective therapies. It is also one of the clearest examples of why AI models require larger, more diverse and longitudinal patient datasets to identify meaningful patterns.
We need to work at speed to deliver successful treatment options, especially as the number of ALS cases is set to rise. There are approximately 140,000 new cases of ALS diagnosed worldwide each year, and this is predicted to increase to 377,000 by 2040.
AI needs data – and lots of it
Harnessing the power of AI to support drug discovery for ALS is key to changing current outlooks. However, AI is only as good as the data we feed it, and the more high-quality and diverse data we can provide, the better.
In the past, obtaining research-relevant data has proven frustratingly difficult in the context of ALS. Data has often been fragmented across different institutions, making interoperability a major challenge.
To make matters even more complicated, each of these datasets can come with its own access requirements, sometimes requiring many months to gain approvals. This makes it incredibly difficult to access, connect and combine the information needed.
This landscape is changing. Principally funded by the MND Association, and bringing together a coalition of charities and data holders from around the world, the Longitude Prize on ALS has convened a dataset that includes whole genome sequencing from more than 9,000 ALS cases and over 3,500 controls.
It also includes comprehensive multi-omics data, comprising epigenomics, transcriptomics and proteomics data from more than 2,000 cases.
Critically, this data is accompanied by relevant clinical information, allowing researchers to connect genetic and molecular findings with real-world patient characteristics.
This feat did not happen overnight. It is the result of tireless international effort and collaboration to overcome the many barriers that have historically prevented such developed access to data.
The prize has now made this data available to 20 teams of international researchers and AI innovators collaborating to achieve one goal: to identify viable drug targets for the development of a successful treatment for ALS.
The critical role of patients
None of this would be possible had patients not agreed to take part in testing and contribute their data. Every patient record that is shared strengthens researchers’ ability to understand the disease and translate that understanding into effective treatments.
Without this collective effort, even the most advanced AI would be unable to deliver.
With that contribution comes a profound responsibility. Patients who choose to share their data are placing trust in researchers, institutions and healthcare systems. They are trusting that their information will be handled securely, ethically and only for the purposes to which they consented.
But there is another responsibility too: ensuring that this data is actually used to advance research, rather than sitting fragmented across disconnected systems and institutional silos.
To understand attitudes towards data sharing, last year we commissioned an international survey of people living with MND/ALS and loved ones of people living with MND/ALS in the UK, USA and Australia.
The results revealed that 74 per cent of respondents would be happy for their, or their loved one’s, biological data to be shared with scientists and researchers to potentially help in the development of new drugs and treatments for MND.
While this represents a clear majority, the ultimate goal should be to enable as many people as possible to contribute their data, ensuring researchers are armed with as much information as possible to understand the disease.
Achieving this depends on building trust with patients and families through transparency, strong ethical safeguards and clear accountability around how data is used. Greater patient data sharing, carried out securely and ethically, is crucial to expanding the depth and quality of information available to researchers across the globe.
Although there is clear willingness from patients to share data, that willingness is not always reflected among data holders. Patient generosity can only drive progress if data can be responsibly accessed and connected across the wider research community.
Locking valuable data behind institutional walls achieves little for patients who already have very few options following diagnosis. Policymakers, governments, researchers and institutions all have a role to play in making data more accessible, more connected and more usable, while maintaining the safeguards and trust that patients rightly expect.
Where do we go next?
The Longitude Prize on ALS is home to one of the largest and most comprehensive collections of ALS patient data of its kind. While this is an important accomplishment, we are determined that progress cannot stop there.
We find ourselves at a pivotal moment. The data exists. AI is advancing at pace. For the first time, we have an opportunity to bring these together in a way that could fundamentally change the trajectory of ALS research and treatment.
But real transformation will not come from data or technology alone. It will come from connection.
That means ensuring that what has already been collected can be safely linked, responsibly shared and meaningfully used across borders, disciplines and institutions. It will also require sustained investment in the infrastructure, systems and collaboration needed to make this possible.
This requires a shift in mindset as much as infrastructure. We need to move from isolated datasets to shared resources, from fragmented efforts to coordinated global action, and from viewing patients as data contributors to recognising them as partners in discovery.
If we can achieve this, the potential is profound, not just for ALS, but for how we approach complex disease more broadly.
The opportunity is here. The question now is how quickly, and how collectively, we choose to act.
