Homology modeling

The use of 3-dimensional (3-D) geometry and sequence information from proteins of known 3-D structure to develop models for proteins whose 3-D structure is unknown. In the first step of homology modeling, search and alignment algorithms are used to find the best sequence overlap of the 'unknown' protein with the sequences of related proteins which have 3-D data. In the second step, the geometry of the 3-D structures is used as a template for generating a 3-D structural model for the regions of high sequence homology in the unknown protein (the conserved regions). Finally, the sections with low homology to known proteins (the variable regions) are modeled using a variety of computational techniques. (http://www.netsci.org/Science/Bioinform/terms.html)