Proteins are the molecules that do everything for us in cells: they allow us to digest food, carry oxygen or allow us to stand up and move our arms. To do so, they need to interact with other molecules. The video above shows a protein in grey which is important for cancer (this particular protein is called MDM2). In normal conditions it interacts with a protein called p53 —the guardian of the genome— and tells the cell to degrade p53, which can lead to cancer. Therapeutic solutions aim at developing drugs that interact with MDM2 better than p53 — taking its place. In this way, p53 is free to signal cancer cells to die.
In the video above you see two small molecules (red and blue) that have been developed as possible therapeutics. The video comes from a computer simulation, similar in idea to the simulations that predict the weather or tornadoes. By looking at how two different molecules interact (bind) with the protein and how often we can tell which one is likely to be a better therapeutic (of course there are many more stages after this in drug discovery). By understanding key interactions between the molecule and the protein we can generate new molecules that might interact better. These computational models can save a lot of time and expense in experiments and provide good leads to work in the lab.