Using DALL-E to Unlock the Potential of Artificial Protein Drugs

DALL-E has sparked the race to create artificial protein drugs

The word \”protein\” is often associated with chicken breasts. However, the molecules themselves are similar to a complex Lego puzzle. The building of a protein begins with a series of amino acids, which are like a bunch of Christmas lights strung together. These amino acids then fold up into 3D structures.

DeepMind, Baker and others made waves with their algorithms that predict the structure of proteins based on amino acid sequence. The predictions had to be mapped down to the atomic scale.

The complexity of designing new proteins is a whole other level. Baker’s Lab took on the challenge this year, using both old-fashioned screening techniques as well as deep learning hallucinations. Both algorithms were extremely effective for decoding natural proteins and creating new ones. However, they were difficult to scale up.


Thanks to DALL-E, the Race to Make Artificial Protein Drugs Is On