Molecular Origami

Proteins make up 50% of the dry mass of living tissues. Scientists know of the existence of about 200 million different proteins. The shape of a protein is determined by the folding of its necklace-like chain of bonded amino acids, using their variable side-chains. Four kinds of interaction (hydrogen bonds, ionic bonds, disulfide bridges and hydrophobic twists) produce the protein's 3-dimensional shape, which determines its activity (e.g. as acting as an enzyme or a hormone). It is possible for biochemists to establish the 3-D shape of a protein but this is a laborious, time-consuming process. Step into the ring, DeepMind, a computer known largely for its success in games against humans such as chess and Go. The computer has now been 'trained', using 70,000 proteins whose shapes are known, to develop algorithms which can predict the 3-D arrangement of any protein (https://www.theguardian.com/technology/2020/nov/30/deepmind-ai-cracks-50-year-old-problem-of-biology-research ). This is effectively solves a challenge for Biology that has been around for 50 years. People are suggesting that the new knowledge produced should enable scientists to tease apart the mechanisms underpining some diseases, design medicines, develop more nutritious crops and even produce enzymes that can break down plastics. This seems to be a better use of Artificial Intelligence than playing games.