Searching Molecular Fingerprint

Fig. 13.6. Results from the third validation study. The -axis represents the Tanimoto similarity score of returned hits with respect to their corresponding query molecule, calculated based on the FCFP4 molecular fingerprints (31). The x-axis are drug molecules in Fig. 13.5. Search hits are color coded by the PGVL reactions (VRXN) where they are originated from.

Among the most widely applied techniques for similarity searching are molecular fingerprints (58, 59). The assumption is... 

An increasing interest of the scientific community in recent years has been shown in the fields of combinatorial chemistry, high-throughput screening, - substructural analysis, and -+ similarity searching, for which several similarity/diversity approaches have been proposed mainly based on - substructure descriptors such as -> molecular fingerprints, which are particularly suitable for informatic treatment. 

Xue, L., Godden, J.W., Stahura, F.L. and Bajorath, J. (2003c) Profile scaling increases the similarity search performance of molecular fingerprints containing numerical descriptors and structural keys./, Chem, Inf, Comput, Sci, 43, 1218-1225. 

D similarity searching The motivation to perform 2D similarity searches follows from the similarity property principle. All compounds having a similarity above a certain threshold are retrieved from the database by 2D similarity searching. In most cases the similarity is calculated by one of the aforementioned similarity measures based on molecular fingerprints. Therefore, in the first step the fingerprint for the query structure is calculated. 

Lounkine, E., Hu, Y., Batista, J., and Bajorath, J. (2009) Relevance of feature combinations for similarity searching using general or activity class-directed molecular fingerprints. Journal of Chemical Information and Modeling, 49 (3), 561-570. 

In the LEVS field, one can broadly distinguish between two basic categories of methods similarity searching on one hand and compound classification on the other [3]. The popular similarity search tools as shown in Figure 11.1 include molecular fingerprints derived from molecular graphs (2D) or conformations (3D) [4, 5], pharmacophore models [6], simplified molecular graph representations [7],... 

For practical selectivity searching, 69 inhibitors with at least 50-fold selectivity for cathepsin K over S and L were selected as reference molecules. The search protocol involved two different in-house-developed LEVS approaches, as shown in Figure 11.9 a compound mapping algorithm termed DynaMAD [38] and a specialized type of molecular fingerprint consisting of compound class characteristic substructures, ACCS-FP [61]. 

The fact that mass, infra-red, NMR and ultra-violet spectra and, to a certain extent, also chromatographic retention data can be considered as molecular fingerprints , forms the basis of most computerised library search systems. Retrieval methods for characteristic chemical data and techniques for the comparison of human fingerprints have similar elements the first step is to clean up the raw data, then in many cases a data reduction is carried out by selection of prominent features. Finally, there is the comparison of unknown and reference data patterns, which, for a useful result, requires a statistical correlation to be established. In this paper no attention will be paid to feature selection. 

It is a well-established fact that conunon sub-structural fragments often tend to share similar biological activity. Molecular similarity deals with finding molecules which have a comparable amount of stmctural similarity [65], This is used to find stmctures that are similar to a molecule with less informatioa Molecular similarity is very handy in dmg designing, because it reduces the amount of animal testing, as the recorded data can be extrapolated. In this chapter, we learn the basic concepts of molecular fingerprints, similarity measures, and the use of molecular fingerprints in similarity search. 

As is well known, the Tanimoto similarity coefficient, which is the most widely used similarity measure, exhibits size-dependent behavior [5, 92-95] that can significantly influence the results of similarity searches. A significant part of the problem can be traced to the terms in the denominator of the Tanimoto function that counts the number of elements that are common to both molecular fingerprints. Thus, when molecules of widely varying sizes are treated, the number of elements in fingerprint... 

Wang and Bajorath [97] have carried out an extensive study based on their earlier work [96]. Both studies used the Tversky similarity function given in Equation 15.5.2, to assess how molecular complexity ( size ) and bit density influence the results of similarity searches based on molecular fingerprints. Generally, but not always, molecular complexity and bit density are closely related, that is, more complex molecules tend to have greater bit densities than less complex molecules. A key element of their study is the construction of bit-density invariant similarity functions that account for the distribution of both 1-bits and 0-bits. The functions are based on weighted combinations of terms of the form given in Equation 15.5.2 or 15.5.3... 

Screens traditionally denote the presence or absence of predefined atom-, bond-, or ring-centered substructural fragments. However, one may also use subgraphs of the molecules to generate a set of molecular fingerprints to use as the screen. The screen search checks each structure for those screens present in the query substructure. For maximum effeaiveness the fragments included should occur independently and with equal frequency in the database. " ... 

---