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

Structural Motifs
Overview of Methods
Motif Search

Structure/135.html">Structural motifs are identified by screening structures for patterns of atoms or residues that appear to be functionally significant. Motif-finding Methods can be broadly classified into several categories:

1. Literature-based. Information regarding residues critical for a specific function is extracted from literature and other experimental data sources. While this “expert” approach yields high-quality motifs whose residues are experimentally proven to be functionally important, it is time-consuming and not amenable to automation.

2. Unsupervised search. A set of structures is analyzed for statistically anomalous regions without making any assumptions about their functional role.

3. Single structures. Structures are examined sequentially. Residues located near a bound Ligand, or those specified in the SITE record of a PDB file, are simply designated as a motif, without attempting to cluster the structures or build a consensus pattern.

4. Positive Examples. Motifs are identified by comparing exclusively true-positive structures (i.e., structures of Proteins that either perform the function of interest or belong to a target structural Classification category). Other structures are not considered.

5. Positive and negative examples. Motifs are selected based on their ability to correctly identify true-positive structures while excluding all others.

To improve the signal-to-noise ratio in identifying functionally critical motifs, sequence conservation in Multiple Sequence Alignments and the spatial proximity of residues within the motif are also frequently taken into account.