Abstract

sponding authors Introduction It has been hypothesized that the total number of different protein folds is nite, and roughly of the order of 1000 (Chothia, 1992; Orengo et al., 1994; Wang, 1998). Once examples of every fold are known, protein structure prediction would reduce to the inverse protein folding problem, which consists in identifying which sequences are compatible with a given fold (Drexler, 1981). This alternative view of structure determination is the basis of the new eld of structural genomics, which aims to deliver structural information about most genomederived protein sequences. While it is not feasible to determine experimentally the structure of every protein, useful models can be obtained by fold recognition and comparative modeling, provided there is a comprehensive library of folds (Kim, 1998). Structural genomics is currently focusing on the construction of such a library, and a gure of 10,000 to 100,000 representative proteins has been proposed (Sali, 1998). W

Citations