Essay --

Literature Review Can comparative modelling techniques successfully model an entire genome? Introduction There is a need of detailed description and understanding the structure and function of many proteins. Although the structure and function of protein is best determined experimentally but it can be predicted by comparative modelling (Sanchez and Sali, 1998). Homology modelling or comparative modelling is used to constructs a three-dimensional model of a protein by comparing its sequence similarity to one or more known structures of protein (Jacobson and Sali, 2004). Comparative modelling of protein structure is relevant to functional annotation of proteins based on structure and consequently enhances the impact of genome sequencing, functional genomics and structural genomics on medicine and biology (John and Sali 2003). The complete genetic information about amino acid sequences of different proteins is only provided us by genome sequencing efforts. We are now challenged with assigning, understanding, controlling, and modifying the functions of various proteins encoded by these genomes. This task is generally simplified by native protein three-dimensional structures. The experimental methods used to determine the three-dimensional structures are X-ray crystallography and nuclear magnetic resonance (NMR) spectroscopy (Jacobson and Sali, 2004). These techniques have significant advances but unfortunately many protein structures are not easily accessible by experiments. The computational methods resolved the huge gap between the number of available sequences of amino acid and experimentally solved protein structures (Xiang, 2006). Over the last two years, in the comprehensive public databases, such as SwissProt/TrEMBL and GenPept... ...e than a factor of two (Vitkup et al., 2001). Alignment errors due to both their impact and frequency are the most important single limitation on comparative modelling. Conclusively, from the genome projects, comparative modelling proficiently increases the value of sequence information while it is not yet possible to accurately model all proteins. The main holdups are the difficulties in detection of weak similarities for sequence structure alignment and for fold recognition and absence of structurally defined members in many families of protein. Although only 400 domain folds out of the total of a few thousand are known so in the next ten years, the structure of most globular folds likely is to be determined. Therefore, comparative modelling possibly will be applicable to most of the domains of globular protein close to the completion of the human genome project.