Two-dimensional substructure searches

Fourth, we have noted that the computational demands of three-dimensional substructure searching are much greater than those for conventional two-dimensional substructure searching. It is possible that recent developments in parallel computer hardware ° may provide a simple way of increasing the efficiency of three-dimensional substructure searching, and there have already been several reports of the use of parallel processing techniques for two-... 

Hagadone, T.R. Molecular substructure similarity searching efficient retrieval in two-dimensional structure databases. 

Hagadone, T. R. (1992) Molecular substructure similarity searching Efficient retrieval in two-dimensional structure databases../. Chem. Inf. Comput. Sci. 32, 515-521. 

As previously discussed, different types of spectra have different data structure and interpretation rules. To deduce the structure from these multiple spectra, we need to choose one of the spectra as the base spectrum and then start our analysis. Typically, the base spectrum should have an explicit correlation between substructure and subspectra. For instance, the NMR spectrum provides explicit structure-spectrum relations. We now select the one-dimensional NMR speetrum as the base spectrum to show one of the structure elucidation strategies. Other spectra, such as infrared or two-dimensional NMR spectra, are used as constraints to reduce the search space. This strategy is outlined in Fig. 19. 

Hagadone, T.R. (1992). Molecular Substructure Similarity Searching Efficient Retrieval in Two-Dimensional Structure Databases. J.Chem.Inf.Comput.ScL,32, 515-521. 

Xu, J. (2003) Two-dimensional structure and substructure searching, in Handbook of Ghemoinformatics, Vol. 2 (ed. J. Gasteiger), Wiley-VCH Verlag GmbH Weinheim, Germany, pp. 868-884. 

Historically, substructure search methods were first developed for dealing with 2-D structures, and later these techniques were either extended to become, or incorporated into, 3-D substructure search approaches. The same is true for maximal common substructure searching methods. Two-dimensional methods are not only useful tools themselves for many applications, but also an important foundation for the understanding and development of the 3-D counterparts in many cases. In this chapter, the SSS methods will be introduced first. Then, attention will be focused toward the discussion on MCSS methodologies. 

Fragment descriptors Both two-dimensional (2-D) connection tables for the molecular structures involved, as well as three-dimensional (3-D) information for these structures may generate a variety of fragment descriptors for substructure search systems (for a review of substructure search systems, see Chen ). Active research continues to be carried out in this field, and the interested reader can peruse the review by Downs for more information. 

A module of the CHEM-X modelling system (see modelling section). Storage and retrieval of two- and three-dimensional structures with substructure-search capability. Available databases include Chapman Hall Dictionary of Drugs (15,000 compounds). Chapman Hall Dictionary of Fine Chemicals (120,000 small organics). Chapman Hall Dictionary of Natural Products (54,000), Derwent Standard Drug File (31,000 biologically active compounds), ChemReact (370,000 reaction types) and others. 

Storage and retrieval of two- and three-dimensional chemical structures with substructure search capability. Available databases include SPRESI (2.2 million compounds), Derwent Drug Index (42,000 drugs), EPA Toxic Substance Control Act (>100,000 compounds), and the Pomona College database (26,000 compounds, 9,000 pXa values, 50,000 P values). 

An important distinaion between ALADDIN and the other programs described in this section is that the user specifies at the time of the search the substructural environment of the atoms and the definition of the geometric objects. This is one key to its flexibility. Another distinction is its close tie to molecular graphics with not only the three-dimensional structure, but also the hit atoms identified for the graphics program. A final distinction is that it has provisions for automatically generating both the two-dimensional and three-dimensional structures of molecules proposed for synthesis. 

Third, there is a need to be able to integrate the basic searching facilities with the more sophisticated routines for three-dimensional structure matching that have been described in this review. ° ° ° An obvious related area is the use of three-dimensional structures to derive descriptors for quantitative structure-aaivity relationships. This could be either an extension of the wide use of two-dimensional fragments for substructural analysis stud-ies 8 8i to three-dimensional fragments, or the automatic generation of data for a prediction of potency, e.g., the recent work of Cramer et al. ° ... 

Different problems for substructure searching on chiral compounds arise in two-dimensional structure databases compared with three-dimensional databases. The two-dimensional case is the worse of the two, because two-dimensional databases... 

---