Last year, the research laboratory Deepmind, owned by Google, published the latest news about AlphaFold, a tool capable of predict protein folding and, with this, collaborate in the disease research all types.
Making these kinds of predictions is very necessary, but also very complicated. Many scientists and companies have tried to develop algorithms based on artificial intelligence to make it. But they are usually not effective enough.
For this reason, there is an inspection that every two years is dedicated to analyzing the latest advances in the subject, classifying them by a scoring system from 0 to 100, called GDT. In 2020, AlphaFold received the highest ratings of any tool ever. For this reason, they have continued with the improvements of this application that, according to what has just been announced, already has the entire catalog of human proteins and also of some species widely used in research. What’s more, will be public, so that any scientist who needs to use it can do so at no cost.
The importance of protein folding for disease research
The protein folding is essential for them to be able to correctly carry out their biological function. The process by which they unfold and lose their three-dimensional structure is known denaturation. It can be something harmless and reversible, such as the reason why heat straightens hair. However, if it occurs in proteins with important functions, it can be behind many pathologies.
Therefore, in disease research it is important to know predict these folds. It can be done experimentally, but it is expensive, takes a long time, and is not possible for some proteins, such as membrane proteins, which are very difficult to crystallize. It took 50 years to find a solution to this.
So, Deepmind, from Google, launched AlphaFold. And now, thanks to its latest advances, published in several Nature studies, it has become a simple tool, available to any scientist who needs it.
Google’s DeepMind to the rescue
DeepMind by Google, and the European Institute of Bioinformatics today announced their joining the European Molecular Biology Laboratory (EMBL) to finally turn the database obtained by AlphaFold into an accessible reality.
In it is the protein folding of 20 very important organisms in disease research. This is the case, for example, of bacteria Escherichia coli, the fruit fly, the zebrafish or the mouse. But there are also around 20,000 proteins expressed by the human genome. That is to say, all the proteins for whose synthesis we have instructions in our DNA.
This tool is in its prime, but it’s only just getting started. As they have explained in a statement, in the coming months they hope to expand coverage to almost all sequenced proteins known to science. This would be about 100 million structures, so it still has a lot to grow.