How Google AlphaFold can help end malaria

It’s been seven years since AlphaGothe artificial intelligence program designed by Google Deepmind achieved what seemed unthinkable: that an algorithm could easily defeat the world champion of the Chinese strategic game Go, considered one of the most complex in the world and in which until then, the machines had not been able to demonstrate as much skill as in other games. The successes of IBM y Deep Bluea computer that already in 1996 proved capable of defeating the world chess champion at that time, the Russian Gary Kasparov.

Subsequently, everything good that Google had learned by training AlphaGo was used as a basis for the development of AlphaFold, a new AI platform designed jointly by DeepMind and Isomorphic Labs and whose objective was to apply this knowledge to environments such as molecular biology or medicineso that it could be useful in research into diseases that until now have no cure.

Its third version (Alpha Fold 3) was launched this year and according to Google, it already allows us to predict the structure and interactions of all the molecules that make up life. That this is possible represents a before and after in the field of scientific research and this was recognized by the Nobel Prize committee a few weeks ago: John Jumper and Demis Hassabis, from Google DeepMind in London, and David Baker, from the University of Washington in Seattle, received the 2024 Nobel Prize in Chemistry for this revolutionary breakthrough.

The truth is that we are already beginning to see the first practical applications of what AlphaFold is capable of doing and that demonstrate that the current possibilities of AI go much further than generating texts and images with ChatGPT. One of the most promising scenarios is development of a 100% effective vaccine to combat malaria.

A vaccine powered by artificial intelligence

Few diseases have a greater impact on the world than malaria. In 2020 alone it killed more than 627,000 people, mostly children under five years old. Although the mosquito that causes the disease is found primarily in Africa, in reality more of the world’s population is susceptible to becoming a future victim of the disease.

Although there are palliative treatments and in reality many people can live with this parasite, the main problem it presents is that its symptoms, which can begin with some fever and headache, make it not easy to detect, and it is misdiagnosed. , it is not treated and when it is done, it is already too late.

When an infected female mosquito bites a human, one of the five types of parasites that can cause malaria has a high chance of entering the bloodstream and subsequently reaching the liver. Here they mature and multiply, being released again throughout the body.

The main difficulty in combating this infection lies in the great capacity of the virus to constantly change your appearance and that of the blood cells they infect, allowing them to avoid the immune system. Historically, this is what has greatly complicated the development of effective medications or vaccines. In fact, the most effective (RTS,S), approved in 2021 as a breakthrough, is only 30% effective and is only capable of attacking the initial phase of liver infection.

The biggest obstacle that researchers face is that unlike the few proteins presented by a virus like SARS-CoV-2 (responsible for COVID-19), the malaria virus has hundreds and even thousands on its surface, therefore which is quite a challenge to decipher its molecular structure.

Historically, determining the structure of just a single protein was a process that could take years, using techniques such as X-ray crystallography or cryo-electron microscopy to often provide an incomplete picture. Multiplying this by a hundred or a thousand was a colossal effort for which resources were never enough.

Now, thanks to its ability to predict the structure of molecules, AlphaFold has tremendously accelerated this process, allowing scientists to obtain accurate models in a matter of days. This has accelerated the development of new vaccines so much that in addition to the two already approved, there are 29 vaccines in active development, involved in a total of 77 clinical trials.

Of these, 15 focus on the preerythrocytic phase, 11 on the blood phase and 5 on the sexual or transmission blocking phase. Just five years ago and without the advance that AI has brought in the development of new medications, the possibility of defeating malaria still seemed as distant as it did in the 90s. And although it is true that AlphaFold is not yet perfect, It is also the first time in many years that a window of hope has opened.