Classification of Protein Function

Annotation of a genome requires assignment of functions to gene products, in most cases on the basis of amino acid sequence alone. The goal of structural genomics projects is to make three dimensional information available, which is invaluable in assessing inferences from amino acid sequences. Nevertheless, prediction of protein function, even if sequence and structure are known, is in many cases a difficult problem. Comparative genomics and patterns of interaction sometimes provide essential clues. Some methods provide reasonable guesses at function, but none is foolproof. An underlying problem is that function is in many cases an ill defined concept. In this article we review the state of the art in function prediction.