The artificial intelligence program ‘AlfaFold’ predicts the structures of more than 350 thousand proteins from the human genome.
Some scientists understand proteins as the essential building blocks for life as we know it. This is so because they build our cells. A new artificial intelligence program articulated by Google raises predict all such structures in the human body. In collaboration with other similar apps, they could decipher cures for certain diseases and come up with more effective drugs to achieve this.
Deciphering the human genome
A new study suggests the possibility that an artificial intelligence program designed by Google identify and catalog the shapes of human proteins. According to the scientists in charge of the experiment, this could be a fundamental advance for modern medicine, since we have only managed to find the structure of a minimal fraction of all that exist.
Called AlphaFold, this software can “predict the structures of 350,000 proteins from humans and other organisms,” according to BBC coverage. The program starts from basic instructions contained in our genomes: the genetic code contained in the nuclei of cells.
In this regard, Demis Hassabis, CEO and co-founder DeepMind (the company that developed the software), highlighted that this could be “[…] the most complete and accurate image of the human proteome to date ”. According to the expert, it could provide the widest catalog of amino acids known to date.
We suggest: What is critical thinking and why is it a more important skill than intelligence
A systematic model
Hassabis is sure that AlfaFold is the “most significant contribution of artificial intelligence to the advancement of scientific knowledge to date ”, as emphasized in a statement. The director assured that the program works with the blockchains built by amino acids, and guesses how different diseases will bend.
In the same way, the program is equipped to mimic a unique 3D structure, as it works in the human organism. But nevertheless, it is not limited to our species. Has managed to model some of the bacteria E. coli, fruit flies, and some small rodents, according to the article published in Nature.
So far, the program has achieved a accurate prediction of folds for 58% of amino acids humans. After months of work and considerable investment to finance the project, it seems that it is finally paying off. No other study technique had achieved systematically determine the structures of all proteins produced by the body. AlphaFold is already capable of doing this.
Japan will use artificial intelligence to help its citizens find a partner and have children
4 signs that your emotional intelligence is high