Using Homology Modeling

Template recognition is the process of finding the most similar sequence. The researcher must choose how to compute similarity. It is possible to run a fast, approximate search of many sequences or a slow, accurate search of a few sequences. Sequences that should be analyzed more carefully are the same protein from a different species, proteins with a similar function or from the same metabolic pathway, or a library of commonly observed substructures if available. 

Backbone generation is the first step in building a three-dimensional model of the protein. First, it is necessary to find structurally conserved regions (SCR) in the backbone. Next, place them in space with an orientation and conformation best matching those of the template. Single amino acid exchanges are assumed not to affect the tertiary structure. This often results in having sections of the model compound that are unconnected. 

The generation of loops is necessary because disconnected regions are often separated by a section where a few amino acids have been inserted or omitted. These are often extra loops that can be determined by several methods. One method is to perform a database search to find a similar loop and then use its geometric structure. Often, other conformation search methods are used. Manual structure building may be necessary in order to find a conformation that connects the segments. Visual inspection of the result is recommended in any case. 

---

A rational approach toward the design of class-I isoform selective HDAC inhibitors was reported recently by Wiest et al. [100]. In order to understand the difference between these class-I isoforms, three-dimensional models of HDACl, HDAC2, HDAC3, and HDAC8 were built using homology modeling. 

Qaude, J. B., Suhre, K., Notredame, C., Qaverie, J. M., and Abergel, C. (2004) CaspR a web server for automated molecular replacement using homology modelling. Nucleic Acids Res. 32, W606-609. 

AG7088 (9.1) was discovered using homology modeling. AG7088 is a weak inhibitor of an important protease for the activity of a coronavirus that was responsible for the 2003 outbreak of severe acute respiratory syndrome (SARS) (Figure 9.3). In the search... 

Ostberg, N., and Kaznessis, Y. (2004) Protegrin structure-activity relationships using homology models of synthetic sequences to determine structural characteristics important for activity. Peptides 26, 197-206. 

Reeves DC, Sayed MF, Chau PL, Price KL, Lummis SC. Prediction of 5-HT(3) receptor agonist-binding residues using homology modeling. Biophys J 2003 84(4) 2338-2344. 

A. G. Pavlovsky, D. F. Ortwine. Effect of species differences on stromelysin-1 (MMP-3) inhibitor potency. An explanation of inhibitor selectivity using homology modeling and chimeric proteins. J Biol Chem. 1999, 274, 24881-24887. 

It is generally considered that a sequence identity below 30% is insufficient to build a useful homology model. Success stories of GPCR homology modeling... 

---