Artificial protein did not have the ability to move and change as to nature, an obstacle now overcome thanks to the use of artificial intelligence (AI), which opens new ways for medicine or agriculture.
Proteins catalyze life by changing form when they interact with other molecules, at that time, the result can be a muscle contraction, the perception of light or a little energy extracted from food.
Scientists have been designing rigid proteins since the 1980s and more recently used to produce medications or antibodies against cancer and inflammation, but despite their importance they have less potential.
However, the design of stable proteins and at the same time dynamic, that is, that they can turn, twist and transform in complicated ways and then return to its original form, requires a calculation power and an AI that did not exist until a few years ago.
The new study published by Science presents the method used by a team headed by the University of California in San Francisco (EU) and is the first step of a path that will take far beyond biomedicine, to agriculture and the environment, according to one of its signatories, Tange Kortemme.
The most important EMULAR proteins for medical uses are those that regulate processes such as metabolism, cell division and other basic vital functions, which facilitate communication within the cells or among them changing shape and returning to the original, such as an ignition and off switch.
You are interested: an AI allows to translate the language of proteins linked to Alzheimer’s
Mobile proteins could be key to treating diseases
The team wanted to devise a design method that could be applied in many situations, for which they focused on creating a mobile part that did what many natural proteins do.
The hope, according to Amy Guo, another of the signatories, is that this movement could be added to static artificial proteins to expand what they can do.
The first step was to give a simple natural protein the ability to move in a new way. Then, they generated a virtual library in thousands of possible ways that the protein could adopt and chose: one that could join calcium and another that did not.
The work accelerated thanks to the Deepmind alphafold2 AI system, which allowed the structure of almost 200 million proteins to predict, almost all acquaintances, and that the team used to make the mobile part twist and capture calcium, and then unscrew to release it.
The researchers tested the model in a computer simulation that showed that the protein “worked exactly” as expected, Guo said in a university statement.
In the medical field, mobile proteins could be used in biosensors that change form in response to disease signals, activating an alert. They could also be used as medicinal proteins adapted to each person’s body chemistry.
Proteins that change could also be designed to decompose plastics or help plants to resist climate stress, such as drought or pests, and could even manufacture metals capable of repairing themselves when they crack.
With EFE information
Do you like to inform yourself for Google News? Follow our showcase to have the best stories
