Search

We examined "descriptor collision" for several chemical fingerprint systems (MDL 320, Daylight, SMDL), and for a 2D-based descriptor set. For large databases (ChemNavigator and WOMBAT), the smallest collision rate remains around 5%. We systematically increase the "descriptor collision" rate (here termed "descriptor confusion"), in order to design a set of "descriptors to mask chemical structures", DMCS. If effective, a DMCS system would not allow third parties to determine the original chemical structures used to derive the DMCS set (i.e., reverse engineering). Using SMDL keys, the "confusion" rate is increased to 45.6% by eliminating those keys that have a low frequency of occurrence in WOMBAT structures. We applied an automated PLS engine, WB-PLS [Olah et al., J. Comput. Aided Mol. Des., 18 (2004) 437], to 1277 series of structures from 948 targets in WOMBAT, in order to validate the biological relevance of the SMDL descriptors as a potential DMCS set. The "reduced set" of SMDL descriptors has a small loss of modeling power (around 20%) compared to the initial descriptor set, while the collision rate is significantly increased. These results indicate that the development of an effective DMCS is possible. If well documented, DMCS systems would encourage private sector data release (e.g., related to water solubility) and directly benefit public sector science.

In-house pharmaceutical collections are no longer sufficient for sampling chemical spaces. As novel bioactive chemotypes are successfully identified by virtual and high -throughput screening, the ability to rapidly sift through large numbers of chemicals prior to acquisition or experiment is required. Strategies for compound selection include some of the following steps: 1.) database assembly ('in silico' inventory); 2a.) structural integrity verification (keep unique structures only); 2b.) limited exploration of alternative chemical representations for the uniques (stereoisomers, tautomers, ionization states); 3.) property and structural filtering (remove unwanted structures); 4.) 3D-structure generation (for virtual screening or 3D-based similarity); 5a.) clustering or statistical design for selection; 5b.) similarity-based selection (if bioactives are known); 5c.) receptor-based selection (if target binding site is known); 6.) add a random subset to the final list.

An automated PLS engine, WB-PLS, was applied to 1632 QSAR series with at least 25 compounds per series extracted from WOMBAT (WOrld of Molecular BioAcTivity). WB-PLS extracts a single Y variable per series, as well as pre-computed X variables from a table. The table contained 2D descriptors, the drug-like MDL 320 keys as implemented in the Mesa A&C Fingerprint module, and in-house generated topological-pharmacophore SMARTS counts and fingerprints. Each descriptor type was treated as a block, with or without scaling. Cross-validation, variable importance on projections (VIP) above 0.8 and q 2⩾0.3 were applied for model significance. Among cross-validation methods, leave-one-in-seven-out (CV7) is a better measure of model significance, compared to leave-one-out (measuring redundancy) and leave-half-out (too restrictive). SMARTS counts overlap with 2D descriptors (having a more quantitative nature), whereas MDL keys overlap with in-house fingerprints (both are more qualitative). The SMARTS counts is the most effective descriptor system, when compared to the other three. At the individual level, size-related descriptors and topological indices (in the 2D property space), and branched SMARTS, aromatic and ring atom types and halogens are found to be most relevant according to the VIP criterion.

The current drug discovery paradigm allocates a short period of time (3–12 months) for the process of lead identification. Thus, medicinal chemists have a rather short amount of time to familiarize themselves with prior art i.e., background information related to the biological target and to chemotypes relevant on the intended, or related targets. Gathering such background information is enabled by chemical databases such as Chemical Abstracts via SciFinder, Beilstein and Spresi, by medicinal chemistry related patent databases such as the MDL Drug Data Report, (MDDR), the World Drug Index, (WDI), Current Patents Fast Alert, and by collections of bioactive compounds such as Comprehensive Medicinal Chemistry, and the Physician Desk Reference, (PDR). As pharmaceutical drug discovery relies on chemogenomics, the average end-user has learned to expect database systems with built-in search engines that seamlessly mine chemical and biological data within an information-rich, integrated resource. The integration process itself requires hierarchical classification schemes, such that knowledge related to target-focused chemical libraries and biological target families can be mined simultaneously. Such tools being under development, we address bioactivity databases in general with focus on bioinformatics and cheminformatics resources. Technical details are omitted for clarity. Readers are encouraged to consult the general references, as well as software documentation. The first section, “Databases and Management Systems”, introduces databases in general. The second part, “Bioactivity Databases: Some Design Guidelines” provides guiding principles for designing and maintaining bioactivity databases. The third section “Biological and Bioactivity Information Databases Examples”, provides a brief overview of currently available bioactivity databases.