Molecular Modeling on the Web

With the worldwide genomic sequencing efforts producing a steady stream of data, the sequences of tens of thousands of proteins can now be determined. Understanding protein function requires knowledge of protein structure. Information on the 3-D structure of a protein can provide insight to the protein’s interactions with other molecules; this information can also be exploited in such endeavors as protein engineering, mutagenesis, and computer-aided drug design. Detailed structural information is available for only 10% of all known protein sequences. Of the remaining 90%, a homologous protein of known structure has been identified for 10%– 25%. For this subset, a rapid in silico approach to 3-D structure is accomplished using homology or comparative modeling. For another 25% of the available protein sequences, structural models can be generated using secondary and tertiary structure prediction methods.