31
FOLD RECOGNITION METHODS
INTRODUCTION
Despite a good qualitative understanding of the forces that shape the folding process, present knowledge is not enough to be used for direct prediction of protein structure from first principles, such as the fundamental equations of physics. A related but easier problem is to recognize which of the known protein folds is likely to be similar to the (unknown) fold of a new protein when only its amino acid sequence is known. This has been variously called an inverse folding problem (find a sequence fitting a structure), threading (since a sequence is being threaded through a known structure), and finally a fold recognition problem. Solution of the fold recognition problem is a necessary prerequisite to the solution of the general folding problem. If we are unable to recognize a structure similar to the correct one, how could we possibly arrive at the correct structure starting for a random one? At the same time, even the partial solution to the fold recognition problem offers the immediate advantage of an efficient and fast structure prediction tool.
Such considerations lead to the development of methods, which attempt to recognize possible structural similarities even in the absence of recognizable sequence similarity. Of course, advances in sequence analysis are constantly changing the threshold between fold recognition and sequence-based homology recognition. Therefore, in this contribution, both sequence and structure/energy-based ...
