AlphaFold: protein structure predictions that will revolutionise medicine

For decades, scientists have struggled to predict the ways in which the many proteins in the human body fold to perform a function. Predicting the shape and function of a protein simply and accurately from the DNA template alone is nearly impossible. A method known as X-ray crystallography is often used to predict the way a protein will fold after its translation and isolation. However, this methodology is complicated, expensive, and slow. Recently, huge strides have been made in employing artificial intelligence to make accurate predictions of protein shape from their DNA templates. AlphaFold is one example of a protein structure prediction software that applies machine learning technology to this classic protein folding problem. With a massively increased speed and reduced cost, it now seems inevitable that the production of therapeutic proteins will increase exponentially in the coming years