结构生物学sturcture biology explore-47.pdfVIP

21 Evolutionary predictions of binding surfaces and interactions Olivier Lichtarge* and Mathew E Sowa† Rapid progress in structural biology and whole-genome Sequence variations and function sequencing technology means that, for many protein It has long been recognized that functional sites undergo families, structural and evolutionary information are readily fewer mutations during evolution than other parts of a available. Recent developments demonstrate how this protein [8,9]. Amino acid differences at functional sites information can be integrated to identify canonical can be described in terms of sequence conservation determinants of protein structure and function. patterns using matrices of position-specific variations, Among these determinants, those residues that are on called profiles [10], or through statistical means, such as protein surfaces are especially likely to form binding sites and hidden Markov models (HMMs) [11,12]. Both methods are the logical choice for further mutational analysis can sensitively detect local sequence motifs, which can be and drug targeting. associated with specific functions and compiled, as, for example, in the INTERPRO database [13]. Global sequence Addresses homologies can also be identified by these methods or *Department of Molecular and Human Genetics, 1 Baylor Plaza, by the related and even more sensitiv