The day we learned that our 15-month-old daughter was one of only 200 people globally diagnosed with a rare genetic brain disorder is a moment I will never forget.

For months, I held onto the hope that everything would be fine.

It was evident that Frankie wasn’t progressing as expected. At her age, like her four older siblings, she should have been crawling and perhaps uttering a few words, but she wasn’t.

I attributed her developmental issues to the meningitis she suffered as an infant. With my medical background, I understood that such an infection could delay development. I convinced myself she would eventually catch up and that our concerns were premature.

Even when specialists gently hinted that there might be underlying issues, I was in denial—was it optimism or fear blinding me? To this day, I don’t know for sure.

So when we received the diagnosis on July 7, 2025, it was devastating: Frankie has DeSanto-Shinawi syndrome (DESSH).

Our consultant was amazing, calmly and compassionately explaining what it meant, where we could get help, the therapies – such as physiotherapy and speech and language therapy that might support Frankie’s development – and the worldwide parent groups who would understand what we were going through.

She explained the genetics. We all carry two copies of the WAC gene, which makes the WAC protein that is crucial in early brain development. In Frankie, one single letter in her DNA had changed – a random misprint leaving her with only one functioning copy.

Half the normal amount of protein with a lifelong impact: learning disabilities, problems walking, talking and eating, bowels not working properly and a high chance of having fits and behavioural difficulties as she gets older. All this means she will need lifelong care.

When it came to medical treatment, though, there was nothing. The gene is known, but there is no gene therapy and no prospect of one any time soon.

Delivering corrective genes into the developing brain is extraordinarily complex, and impossible with today’s technology. For a condition as rare as hers, it’s also not commercially worth it for drug companies to look for gene therapies, as so few children could benefit.

When you have a child who is unwell, a diagnosis can be a relief, the start of the process of making them better. I’ve spent my career in emergency medicine, where I know how to treat my patients – you may not always win, but there is certainty in what needs to be done.

But with Frankie, my daughter and the most precious patient I’ve ever felt responsible for, I didn’t have a clue what to do. I was facing a waiting game to see what would happen to her without me being able to fix it.

At home, we did everything we could. We poured into Frankie more love, structure and support than I thought possible, helped by her amazing medical team and her utterly brilliant nursery.

We celebrated small gains, like when she learnt to blow a kiss, and we adjusted our expectations. But as a doctor, I struggled to accept that therapy and love were the only tools available to her.

I read everything I could find about the WAC protein. I went from knowing nothing about it to being able to recite the key papers from memory.

I studied zebrafish and genetically modified mice to understand what happens when specific genes are switched off – to understand which systems in the brain were affected, and which ones might potentially be adjusted or treated with drugs.

I also connected with the DeSanto-Shinawi community, including The DESSH Foundation, a support group founded by Caitlin Piccirillo, whose child has DESSH. She organises an annual research clinic, run by Professor Marwan Shinawi, who first identified the condition in 2015. It assesses affected children and gathers data and biological samples, such as blood and skin biopsies, to better understand the condition and support future research.

And the more I read, the more I realised that the biology was not as mysterious as I thought.

Genetic diseases are essentially down to problems with the instructions that genes send to cells, affecting the production of specific proteins, such as the WAC protein in Frankie’s case. While these are rare diseases – individually, each condition affects a few children and adults – collectively, they affect hundreds of thousands in the UK alone.

Then I came across the work of Matthew Might, whose efforts are surely going to change so many of these lives.

Fourteen years before my own family’s diagnostic bolt out of the blue, Matthew had sat through a version of the same conversation about a different ultra-rare condition.

His son Bertrand, then five, had developed differently. There were years of appointments and inconclusive tests, before whole-gene sequencing – where your entire genetic code is read – finally identified a mutation in a gene called NGLY1, which produces a protein involved in the cell’s waste-disposal system, helping to clear damaged proteins.

At the time, only a tiny number of children anywhere in the world were known to carry similar changes.

There was no named condition then, no clinic, no research pathway, no pharmaceutical interest. Just a genetic report and a long list of unknowns.

Matthew wasn’t a doctor. He was a computer scientist working at the University of Alabama in the US, an expert in machine learning and artificial intelligence long before AI was understood by anyone outside of research institutes and cutting-edge tech companies.

Instead of accepting that the identification of a mutation was the end of the story, he treated it as the beginning.

He began examining the data – searching databases, analysing genetic networks, reaching out across the internet to find other families whose children carried similar mutations.

Slowly, painstakingly, isolated cases were connected and what had been scattered anomalies began to form a recognisable condition: NGLY1 deficiency.

Like DESSH, it’s super rare, thought to affect just one in five million people.

He recognised that it would not be possible to recruit thousands of patients for research and trials.

So he used his skills to analyse millions of data points about medicines we already use safely to treat other conditions, to identify those that affect the biological systems involved in a genetic disease.

In other words, this is not about inventing a brand new drug, which can take decades, but working out whether existing licensed medicines could gently adjust these disrupted systems and offer hope in treating rare conditions.

And AI is key – analysing all the data out there in the world and bringing it together.

This means the question could move from ‘can we fix the gene?’ – which is often impossible – to ‘can we change the biological pathway, using AI-discovered repurposed drugs – a different, unexpected drug for each of the thousands of rare genetic conditions?’.

That way of thinking now underpins the National Institutes of Health Biomedical Data Translator in the US – AI software, in which Matthew has been a key player, that’s free for researchers to use.

Matthew used his AI modelling to help find drugs for his son that improved his quality of life and helped him live far longer than doctors predicted.

Sadly, there was no miracle cure. In 2020, Bertrand died from complications of his condition. But his father has shown that AI’s power to generate potential novel treatments could be transformative and, crucially, offers hope for parents like me.

It is also, for once, some good news about the power and potential of AI.

This strategy’s ability to treat rare genetic diseases is so novel and so new that many of us in medicine (myself included) couldn’t see the significance of what was happening in plain sight – or how much potential it could offer children like my daughter.

Luckily, not everybody was as blind as I was.

Laura Lambert, who has a PhD in genetics, genomics and bioinformatics, had previously worked alongside Matthew at the University of Alabama, with a particular focus on proteins in rare genetic mutations.

When she moved to the Mayo Clinic in Rochester, Minnesota, a few years ago, she teamed up with Dr Whitney Thompson, a specialist in both genetics and neonatology, and together they began exploring whether the Translator concept could move from theoretical to real-world application – in other words, using AI analysis to generate plausible drug candidates for a specific child.

A little girl named Jorie, born a few months before Frankie, became their first patient.

As it happened, she was also diagnosed with DESSH. Her parents had the same conversation I had, including the careful explanation of the gene and the fact that there was no established treatment, beyond therapy and support.

But alongside that conversation, her parents were told about this new computational approach, and that the team at the Mayo Clinic had identified licensed drugs that might increase the amount of WAC protein Jorie’s body produced.

One of those drugs was a medication widely used in paediatric neurology to treat epilepsy, with a good safety profile in children.

Before prescribing anything, the team took a sample of Jorie’s blood and skin biopsies.

When the drug was added to her cells in the laboratory, WAC protein levels increased, approaching those seen in cells without the mutation.

After careful discussion with her parents, balancing risk against the absence of medical alternatives, her team started Jorie on the medication just before her second birthday.

Jorie’s mum later described what followed as if ‘her lights had been turned on’.

Jorie began gaining milestones she had previously missed, such as her speech and understanding, accelerating in a way that was hard to ignore.

It wasn’t magic and it wasn’t certainty. Of course, the benefits Jorie was seeing could have been coincidence. They could have reflected her normal developmental trajectory.

Indeed, the way DESSH affects a child is hugely variable: some progress steadily, some plateau, some regress.

And the experience of just one patient is not proof. It is a signal.

Jorie’s mother happened to work for an American news channel, and she made a YouTube video about her daughter’s journey. It was shared on the DESSH WhatsApp groups that Caitlin had organised.

I watched it late one evening, about two weeks after Frankie’s diagnosis and, as I did, the penny dropped. This wasn’t just a hopeful anecdote, this was ground-breaking science.

If it were proven the improvement was real – that it was drug-driven rather than coincidence – then this wouldn’t just change how we treat one condition, the model could be applied to thousands of rare genetic diseases.

My despair began to turn into something else – a determination to help.

Seven months ago, I was fortunate enough to meet Laura, Whitney and the other scientists involved in this work. As I spoke to them, it became clear that the potential is enormous.

These trials are comparatively inexpensive, as the drugs already exist and their safety profiles are known. You do not need to invent a molecule or run decade-long toxicity studies. But funding is crucial. However, there is no commercial incentive to fund repurposed drugs.

Pharmaceutical companies cannot make profits from off-patent medications, particularly when each disease affects only a few hundred children worldwide.

So the trials aren’t going to happen. And without funding, children like my daughter will not benefit from the potential treatment avenues that these AI tools are creating.

That is why, out of the upset of Frankie’s diagnosis, Rare People – The Research Charity, was born.

As well as helping to found it, I am one of the trustees and our mission is simple: to raise funds to support high-impact, properly designed clinical trials of repurposed, AI-identified drugs to try to treat rare genetic neurodevelopmental conditions.

The first funding priority is to support a trial of children (and, in time, adults) with DESSH.

A study has been approved at the Mayo Clinic and we are working through the process of how children elsewhere (including in the UK) can take part. Frankie cannot simply be given the drug that Jorie had, because one child’s experience is not proof a treatment works. A formal clinical trial is needed to ensure the dose is right, safety is monitored, and any benefit is real rather than just coincidental.

Last week, National Rare Disease Day, organised by the Genetic Alliance UK (Rare People is now a member), once again highlighted that people with rare diseases are often neglected.

Not because no one cares, but because the medical system is not built for small numbers.

Rare People has been set up to close that gap by funding serious, high-quality research in this field across multiple rare conditions.

The founders and trustees are currently meeting the charity’s core operational costs separately, so that funds raised from the public can go directly to supporting research.

Decisions about which projects to fund are made by our Medical and Parental Advisory Board, which includes clinicians, scientists and parents of children with rare conditions.

I have spent my entire career helping others but, for the first time in my life, I am asking for help. Not just for Frankie, but for hundreds of thousands of children to give them more than therapy and hope.

If you believe that rare children deserve the same scientific ambition as those with common diseases, please support us.

Frankie is a perfect, wonderful, loving, joyful little girl. I wouldn’t change her for the world. But I do want her to have the same hopes, the same opportunities, the same freedom to dream about her future that my other children will have as they grow older.