Alignment methods optimal

All alignment methods require some quantitative measure or fitness function, to assess the degree of overlap between the ligands being aligned and to monitor the progression of that optimization. This is most often manifested as a molecular similarity score or alignment index [22]. 

Alignment methods are also known to combine different optimization methods, such as a genetic algorithm and a directed tweak method [36]. 

K. Ogata, M. Ohya, H. Umeyama. Amino acid similarity matrix for homology modeling derived from structural alignment and optimized by the Monte Carlo method. J Mol Graph Model. 1998, 16, 178-189, 254. 

Most sequence alignment methods seek to optimize the criterion of similarity. There are two approaches of sequence alignments a global alignment compares similarity across the full stretch of sequences while a local alignment searches for regions of similarity in parts of the sequences. 

The approaches that are frequently used to construct structural alignments rely on the two classes of representation described previously. The first type comprise structural alignment methods that focus on shape, lattice, or three-dimensional grid representations MSA, HASL, and CoMFA (Figure 2). Since the conformational changes are not taken into account in the optimization of similarity, an indirect validation of the alignment can be provided by the statistical robustness of the derived quantitative model. In contrast, the structural... 

Lcularly popular [Taylor and Orengo 1989], The method is so named because it uses two amic programming steps. In the first step, it is necessary to determine the score for each of residues, one from each structure. These scores are used to fill a rectangular H matrix, irich dynamic programming is applied to determine the optimal alignment. 

The Surflex-Sim method operates significantly differently [104]. Each of the molecules is surrounded by a set of observer points that characterizes the local character of the surface and potential interactions. Two similar molecules will have a common subset of comparable observer points. A optimal alignment occurs when the differences in pharmacophore character and molecular surface inferred from the observer points are minimized between two molecules. To speed up the algorithm, large molecules can be fragmented into parts which are then compared, and then tested for consistency. This feature also makes the method capable of identifying alignments when one molecule is much smaller than the other. 

---