From scratch, an artificial intelligence system creates protein that destroys bacteria
To achieve this feat, an artificial intelligence system called ProGen was used to design millions of new proteins, which subsequently created a small sample to test their performance, in real life and with very satisfactory results.
ProGen learned to generate new proteins by learning the amino acid combinations of 280 million existing proteins from more than 19,000 families.
One hundred physical proteins were created, of which 66 were tested in chemical reactions similar to the natural bacteria-killing proteins in egg whites and saliva, suggesting that they can also kill bacteria.
This advance is fundamental when it comes to making new medicines. In addition, it could help to reduce costs in the process of developing new drugs. We are talking about the fact that medicines go through several processes to reach the patient's hands, from research, development, tests and studies in animals and humans, in addition to the processes carried out by the regulatory agencies to obtain approval, it is not only a monetary resource, the delay that could take between tests means that the patient receives a treatment much later than what is needed to fight his disease, everything points out that using ProGen is more successful than studies carried out by humans, because it is precisely the combination of proteins that usually takes longer.
The research team's proposal is to extrapolate to new protein spaces that have not yet been exploited, and to offer new unique possibilities beyond the natural or patented proteins that we know today. This technology is poised, according to the team, to operate on multiple different modalities of enzymes, antibodies, gene editors and peptides. Hence the promise of effortlessly managing the task of optimizing multiple attributes simultaneously, reducing costs and minimizing time in new product development.
The study was conducted by Ali Madani and colleagues at Salesforce Research in California. James Fraser of the University of California San Francisco was also part of the team.
Read more at https://madani.ai/ & https://www.profluent.bio/
