AlphaFold predicts novel human proteins with knots

The fact that proteins can have their chain formed in a knot is known for almost 30 years. However, as they are not common, only a fraction of such proteins is available in the Protein Data Bank. It was not possible to assess their importance and versatility up until now because we did not have access to the whole proteome of an organism, let alone a human one. The arrival of efficient machine learning methods for protein structure prediction, such as AlphaFold and RoseTTaFold, changed that. We analyzed all proteins from the human proteome (over 20,000) determined with AlphaFold in search for knots and found them in less than 2% of the structures. Using a variety of methods, including homolog search, clustering, quality assessment, and visual inspection, we determined the nature of each of the knotted structures and classified it as either knotted, potentially knotted, or an artifact, and deposited all of them in a database available at: . Overall, we found 51 credible knotted proteins (0.2% of human proteome). The set of potentially knotted structures includes a new complex type of a knot not reported in proteins yet. That knot type, denoted 6(3) in mathematical notation, would necessitate a more complex folding path than any knotted protein characterized to date.

Automated summary

The fact that proteins can form knots in their chain has been known for almost 30 years. However, only a small fraction of these proteins are available in the Protein Data Bank due to their rarity. Recent advancements in machine learning methods have allowed us to analyze the entire proteome of a human organism, leading to the discovery of 51 credible knotted proteins (0.2% of human proteome). This research also unveiled a new type of knot, denoted as 6(3), which requires a more complex folding path than any previously reported knotted protein.