Function class fingerprints

Inspired by this work (24), we used Bayesian methods (24) with molecular function class fingerprints of maximum diameter 6 (53) to identify substructures that were shown to be important in recent TB screening datasets (21-23). Bayesian models were built with the previously described Molecular Libraries Small Molecule Repository (MLSMR) 220,463 library (4,096 active compounds) (15) and dose-response data using 2,273 molecules (475 active compounds). In addition, these models were tested (15) with the National Institute of Allergy and Infectious Diseases (NIAID) data and GVK Biosciences (Hyderabad, India) datasets used by Prathipati et al. (24). We have validated the models with compounds left out of the original models, in some cases showing up to tenfold enrichments in finding active compounds in the top-ranked 600 molecules (22). 

Many classes of enzymes have their own specific sequence fingerprint. Often a sequence is determined that shows no significant homology to any other sequence, except for a very small pattern of residues. Using a library of function-specific patterns one can in many cases still determine the functional class of the protein, despite the foot that no structural information is available. Hie Prosite [92] sequence pattern database is the most extensive depository of function-determining sequence patterns. 

In the interpretation of infrared spectra of chemicals related to the CWC, it should be remembered that for some chemicals, for example, nerve agents, a structure could be proposed easily and accurately. For some other types of chemicals, for example, mustards and other vesicants, only the chemical class can be proposed. This is due to the fact that some functional groups do not have good structure specific group frequencies. Such spectra can only be used as a fingerprint for the chemical. 

Qualitative tests may be performed by selective chemical reactions or with the use of instrumentation. The formation of a white precipitate when adding a solution of silver nitrate to a dissolved sample indicates the presence of chloride. Certain chemical reactions will produce colors to indicate the presence of classes of organic compounds, for example, ketones. Infrared spectra will give fingerprints of organic compounds or their functional groups. 

Wang and Bajorath [84] have developed a position-dependent scheme that weights individual positions in a vector-based similarity function but not the positions of the individual fingerprints. The weights are derived for specific compound classes associated with various biological activities. In addition to their use in vector-based methods, weighting factors have also been employed in methods based on continuous functions [80, 81]. 

However, the question remains as to how the independence of two similarity measures can be established in analogy to the judging example described earlier. One possible approach is to determine the relationship of the representations to each, but it is not immediately obvious how to accomplish this since the mathematical forms of different representations can be quite varied (from molecular fingerprints to property vectors to 3D electron density or related distribution functions). While it may not be possible to solve this problem in general, it may be possible to solve a more limited subproblem by, for example, assessing the relationship of different representations within the same general class such as molecular fingerprints. 

The first step in this fingerprinting process is the separation of component materials from the composite sample. Once separated, the functional groups or class of the component materials may be identified by FTIR spectrometry. This separation may be affected by either physical or chemical means. 

---