Protein Structure and Function. Application of Bioinformatics Methods - John Rigden 2014

Protein Function Prediction Based on Surface Properties
Surface Properties
Surface Conservation

The pseudo-random process of divergent evolution implies that most residues in Proteins with conserved Functions will be conserved, as they are essential either for the protein's Structure or its function. The residue conservation profile is generated by analyzing a Multiple Sequence Alignment of a protein family—using, for example, the Scorecons method (Valdar 2002)—and is subsequently mapped onto the protein surface. This conservation map (which carries no physical or chemical information) allows for the identification of conserved regions on the protein surface. Such a map can be particularly useful in determining the binding site and active center for a given protein family.

Typically, only a small fraction of a protein surface is involved in its function. Only a few residues are key to catalysis within the Active Site. Even for Protein Functions that engage a large surface area, such as Protein-Protein Interactions, only a few residues make a significant contribution to The formation of a stable complex (see Section 7.5.2). These “functional residues” are quite frequently conserved among evolutionarily related proteins with similar functions. Thus, comparing conserved residues on the protein surface provides insight into its functions. Interpreting conservation is often challenging because, unless the proteins are clear orthologs, a mutation may occur that completely alters the protein's function, resulting in a mixture of paralogs and orthologs within the analyzed protein family. Currently, Methods exist only to exclude paralogs from a multiple sequence alignment (O’Brien et al. 2005); therefore, caution must be exercised when comparing sequences with 30% identity, as in such cases only half of the binding sites are predicted correctly (Devos and Valencia 2000). However, high sequence similarity does not necessarily imply functional similarity (Todd et al. 2002). The TIM barrel protein family (Triosephosphate isomerase barrel) demonstrates that the exact same structure can encompass a vast diversity of functions, as discussed in more detail in Chapter 6.

Another challenge is determining which of the conserved residues is the most functionally significant. The Evolutionary Trace method (Lichtarge et al. 1996) uses a Phylogenetic Tree to approximate the evolutionary process, thereby yielding a set of sequences that share some degree of similarity with the query sequence. The hierarchical branches of this tree represent groups of sequences with conserved functions, where the closest groups at the branch tips share the most similar functions. Residues that are conserved at the same position in sequences of one branch, but are less represented in other branches, are referred to as “evolutionary trace” residues. These residues serve as a defining feature of the branch and are presumed to be essential for the function of proteins within that branch. The ConSurf server (Pupko et al. 2002) employs a Modification of the basic evolutionary trace method to color the Cell/13.html">Protein Structure, highlighting the spatial distribution of conserved residues on the protein surface.