DeepMind, a British subsidiary of Alphabet, has announced that it has solved a major challenge in biology known as the “protein folding problem”. The company of artificial intelligence had to develop a method by which the structure of proteins could be predicted solely on the basis of their amino acid sequence. Using a high-tech method called AlphaFold, the company says it will sequence any protein that scientists know. And it has already published the structure of 350,000 proteins in a searchable database.
The MIT Technology Review reported on DeepMind’s latest use of AlphaFold. The company has very much created a searchable database typical of Google that anyone can use online. The AlphaFold DB tool is already working and people can search the entire human proteome. As well as the proteomes of 20 other scientifically relevant biological organisms, such as yeast, fruit flies, and mice.
AlphaFold is able to structure the various proteins that make up the body’s proteome. Using machine learning, the full set of proteins that make up a living organism. DeepMind has trained artificial intelligence on roughly 170,000 protein structures from a large database. The training allowed artificial intelligence to learn which amino acid sequences lead to which types of proteins. Thanks to the patterns obtained from the data, AlphaFold is now able to identify new protein structures only by knowing their amino acid sequences.
As for AlphaFold DB itself, it will not be used too often by the general public, although search tool is easy to use. Users can find proteins by organization name, which is a simple task. But once you find a protein of your choice, such as a molecule called “DnaJ homolog subfamily B member 4” inside a human liver cell, you can’t do much else, just look, thanks to AlphaFold DB, which offers a 3D representation of each protein structure
As is well known, the AlphaFold protein structure database provides the most complete and accurate picture of the human proteome, doubling knowledge of the structure of human proteins. A proteome is a set of proteins (and peptides) expressed by a genome, cell, tissue, or organism at a given time under specified conditions. Proteins are organic compounds made up of amino acids with specific biological functions. The amino acid sequence is encoded by the nucleotide sequence of the genes. The number of genes encoding the protein is estimated at about twenty thousand.
Hardware, software, tests, curiosities and colorful news from the IT world by clicking here
