Leave a piece of string in your pocket, and the chances are it'll be tangled and knotted when you take it out. And yet pick a string of amino acids, and the chances are this protein will not be knotted. Out of the 53,000 proteins in the Protein Data Bank (PDB), only 300 are knotted, and most of these are the same protein in different forms. In fact, it's possible to reduce the total number of different knotted proteins to 17.

Why are knotted proteins so rare?

Today, Martin Lundgren at Uppsala University and his mate Antti Niemi make a tentative suggestion. They point out that biologically active proteins are compact objects, presumably because there is some kind of advantage in being small.

They then calculate the effect that a knot has on the size of a protein and find that knots tend to increase the size of short proteins with fewer than 400 central carbons in their structure. But for proteins with between 400 and 500 central carbon atoms, knots seem to make them more compact.

"The reason why knots are so rare in PDB data would then be partially explained by the fact that until now, the structure of only relatively short proteins have been reliably resolved," they say.

So the implication is that knotted proteins aren't rare at all--only undiscovered. That sounds like a challenge.

Ref: arxiv.org/abs/0906.4950: Knots and Swelling in Protein Folding