On meeting Deepmind

By 13th February 2020 No Comments

Earlier this month I was invited to speak at Deepmind. It all came about in a roundabout way, because I was having coffee with a friend. She’s the best kind of friend. I don’t see her very often, but she’s that kind of person who can see the depth, acknowledge that things are hard and my heart hurts, but also move on and talk about lighter things, and things that are happening in her life. She doesn’t shy away from the hard and doesn’t let the hard weigh down everything else and I appreciate that.

She was talking about her work, and how for some, it’s hard to understand the real world effect of what they’re doing, which makes it hard to understand their purpose. Their why.

They’re all up in the code and the science, she said. When I asked what it was they were working on, she said ‘protein-folding’

I took a moment, just a beat, before I pointed at Mikaere. His entire disorder is because of protein folding.

Truth: it’s actually about protein misfolding.


Science sidenote: Your genes create little recipes, and your body, following these recipes creates protein chains of ingredients (amino acids). This protein chain folds in a particular way, and off it goes into your body to be used.

But, Mikaere has two little missense mutations. Which means that in his recipe, there are two wrong ingredients. These wrong ingredients go into the protein chain and cause the way it folds to be different – it misfolds – meaning it can’t be used.

Kind of like, if you have a recipe for cake, and instead of sugar you put in salt. The cake can’t do its job of being delicious.

Mikaere’s mutations mean in his little GLDC protein recipe, seriene is substituted with leucine and glutamine is substituted with lysine. So, when his body builds out this recipe chain of amino acids, the protein misfolds, and the protein can’t do it’s job in the glycine cleavage system.

Protein misfolding. It’s a thing.


When I said it, I could see something had clicked with my friend. Suddenly her day to day of high level theoretical science had a very real world meaning. AlphaFold – an AI system used to predict protein structure and how they fold – meant so much more with what it might be able to do in the future.

Which is how I found myself earlier this week at Deepmind. I’d organised to have Prof. Nick Greene (from UCL, he heads up one of the very few NKH Research teams) be there too, because I was positive there would be some intense science related interest I wouldn’t be able to talk to. (I was right).

Deepmind, they are so lovely. I was in a room with some very very smart people, and while our story was more the reality of NKH and protein misfolding, when Prof. Nick spoke about the Science (capital s, lots of big words), you could see the cogs turning. There were a lot of interesting science questions (which were well over my head) but that so many people were thinking about the multiple aspects of glycine related metabolism, that was exciting for me.

Even more exciting was hearing about the interesting things Prof. Nick is working on. YOU GUYS, the research that we’re fundraising has such a huge effect, and while I’m not privy to share (Prof. Nick is writing the paper now, so hopefully it won’t be too long before it’s out) I was really really excited about the real world applications, and what that might mean for kids with NKH in the future. Things are HAPPENING, Prof. Nick is learning so much about NKH and glycine and just – I have so so much hope. So much hope!

I was grateful to share our story with Deepmind. I’d gone in with the intention to share our story (because we want Mikaere to be known) but I hope I touched a chord. I’m hopeful that one day AlphaFold will be used to aid scientific discovery in rare disorders like NKH.

Also, sometimes it’s just a privilege to be able to share with people who have influence over things like this. I think AlphaFold is going to be huge for the rare disease science community.

So yes. Thank you Deepmind for having us!


Leave a Reply