Proteins which specifically bind to a target molecule have quickly gained the interest of researchers and the pharmaceutical industry alike. They can be used as markers for different cell structures or molecules, and if the target is chosen correctly, they can affect the progress of a disease. But an important gap remained: how to de novo design binders towards a chosen target? Rapid advances in artificial intelligence tailored to this task made this goal reachable. We chose microtubules as the target and used three different programs to design 27 potential binder candidates. We expressed them fused with a fluorescent protein. After isolation and chromatography, the binding was analysed using total internal reflection fluorescence (TIRF) microscopy. In this research, we show that currently available AI driven tools allow us to design novel proteins binding to a chosen target.
|
