AI as a tool for science - EMBL-EBI and AlphaFold - Google DeepMind

The Power of Data: How AlphaFold is Revolutionizing Biology

Scientists build on the shoulders of giants, and indeed, those shoulders are often data. In fact, over the last 20 years, there has been a significant change in the way we generate data, and this shift has had a profound impact on various fields, including biology. The advent of AI systems like AlphaFold, developed by DeepMind, is a testament to this trend.

AlphaFold is an AI system that can take a protein amino acid sequence and create highly accurate 3D models of that protein. This technology has the potential to generate millions of protein structures, which is akin to having a different microscope for molecules of life. When AlphaFold was launched last July, it made 365,000 protein predictions. However, this number grew rapidly, reaching about 1 million, and now it's two orders of magnitude more.

One of the most significant challenges in working with such massive amounts of data is scaling infrastructure to handle it. This requires getting creative with how we manage our databases. Databases are a bit like plants; they need pruning, regenerating, and fertilizing to continue growing. In this case, we're leveraging the Google Cloud infrastructure to ensure that our system can scale well and handle the enormous amount of data generated by AlphaFold.

The impact of AlphaFold on biology cannot be overstated. It's not just disrupting traditional fields like structural biologists; it's also empowering them to solve structures that were previously impossible. With AlphaFold, scientists can instantly view their favorite protein, which is truly fantastic. The potential for discovery and innovation in this field is vast, and it's an exciting time for biology.

The collaboration between AlphaFold developers and researchers has been a natural one. By taking DeepMind's predicted models and making them available to everyone, we're democratizing access to this powerful technology. This collaboration also highlights the importance of data-driven research in advancing our understanding of biology and medicine. As we continue to explore the capabilities of AlphaFold, we can expect significant breakthroughs in various fields, including structural biology and medicine.

The scale of the data generated by AlphaFold is truly staggering. With millions of protein structures at our disposal, we're on the cusp of a revolution in biology. This technology has the potential to change the face of biology once again, empowering researchers to tackle problems that were previously unsolvable. As we move forward, it's exciting to think about the possibilities and discoveries that await us with this powerful tool.

"WEBVTTKind: captionsLanguage: enscientists build on the shoulders of giants in fact those shoulders most often are data she has completely changed in the last 20 years and the big change has been that we can generate massive amounts of data Alpha Ford is an AI system developed by deepmind and what it can do is take a protein amino acid sequence and it can create these very accurate 3D models of that protein when one of the predictions it can generate millions of protein structures it's like having a different microscope molecules of life when we launched Alpha for DB last July it was 365 000 protein predictions and then it grew to about 1 million and now it's two orders of magnitude more emblem ebi the home in your molecular data and so it was a very natural joining up with deepmind to take their predicted models and make them available for everybody so one of the challenges with the data that we have from from this collaboration is is the size of the data sets and because of that we had to get creative in in the infrastructure and especially how it will scale and databases are a little bit like plants they need pruning they need regenerating they need fertilizing in some way we are taking advantage of the Google Cloud infrastructure we had to make sure that we really have something that cares well that can continue to grow having all these millions of structures will change the face of biology again it's disrupting both the fields of structural biologists but it's also empowering them to solve structures that they couldn't before medicine and I just get your favorite protein and look at it instantly I mean it's just fantastic foreignscientists build on the shoulders of giants in fact those shoulders most often are data she has completely changed in the last 20 years and the big change has been that we can generate massive amounts of data Alpha Ford is an AI system developed by deepmind and what it can do is take a protein amino acid sequence and it can create these very accurate 3D models of that protein when one of the predictions it can generate millions of protein structures it's like having a different microscope molecules of life when we launched Alpha for DB last July it was 365 000 protein predictions and then it grew to about 1 million and now it's two orders of magnitude more emblem ebi the home in your molecular data and so it was a very natural joining up with deepmind to take their predicted models and make them available for everybody so one of the challenges with the data that we have from from this collaboration is is the size of the data sets and because of that we had to get creative in in the infrastructure and especially how it will scale and databases are a little bit like plants they need pruning they need regenerating they need fertilizing in some way we are taking advantage of the Google Cloud infrastructure we had to make sure that we really have something that cares well that can continue to grow having all these millions of structures will change the face of biology again it's disrupting both the fields of structural biologists but it's also empowering them to solve structures that they couldn't before medicine and I just get your favorite protein and look at it instantly I mean it's just fantastic foreignscientists build on the shoulders of giants in fact those shoulders most often are data she has completely changed in the last 20 years and the big change has been that we can generate massive amounts of data Alpha Ford is an AI system developed by deepmind and what it can do is take a protein amino acid sequence and it can create these very accurate 3D models of that protein when one of the predictions it can generate millions of protein structures it's like having a different microscope molecules of life when we launched Alpha for DB last July it was 365 000 protein predictions and then it grew to about 1 million and now it's two orders of magnitude more emblem ebi the home in your molecular data and so it was a very natural joining up with deepmind to take their predicted models and make them available for everybody so one of the challenges with the data that we have from from this collaboration is is the size of the data sets and because of that we had to get creative in in the infrastructure and especially how it will scale and databases are a little bit like plants they need pruning they need regenerating they need fertilizing in some way we are taking advantage of the Google Cloud infrastructure we had to make sure that we really have something that cares well that can continue to grow having all these millions of structures will change the face of biology again it's disrupting both the fields of structural biologists but it's also empowering them to solve structures that they couldn't before medicine and I just get your favorite protein and look at it instantly I mean it's just fantastic foreign\n"