2 ms·

While the article is correct that knowing the protein structures in itself is not that interesting, it's a prerequisite step to predicting interactions between

While the article is correct that knowing the protein structures in itself is not that interesting, it's a prerequisite step to predicting interactions between proteins, which is super interesting for drug discovery. What's encouraging is the rate of progress, not what has already been done.

ramraj07 4y ago
It’s not necessarily a pre-requisite: many of the more “exciting” drug platforms like Antibodies (which comprise like 90% of approvals nowadays anyways) in principle DONT NEED any structure info a priori. You don’t need to know a proteins structure to generate antibodies, just the protein itself to either inject into an animal or pass through a phage display column. Now typically drug companies will of course only work with proteins whose structure is known. Please note that the top drug companies have some of the best crystallography departments and they will most definitely get the structures done experimentally (if it’s possible). AlphaFold will likely help with that for sure. But it’s just one of those things that’s just a nice to have. If a protein is truly important and you cannot tell it’s structure you can still go ahead and make an antibody drug against it and Even characterize it using mutagenesis analyses.