Atom-specific descriptor

Beyond similarity-based applications, machine learning techniques may pick the specific descriptor elements that appear to correlate with the observed activity trends throughout a training set. Unlike in overlay models, where there is an obvious link between pharmacophore spheres or fields in space and their source atoms, the actual pairs (triplets, etc) of atoms in molecules that incarnate the picked descriptor elements must be first established, to gain any potential insights into the binding mechanisms. 

By restricting RDF descriptors to certain atom types an atom-specific RDF descriptor. 

A structural feature is defined by nine numbers. The first four numbers (nl-n4) serve to identify the specific atom type, atom pair, or atom environment by means of a predefined set of properties, while the remaining five numbers (n5-n9) determine which bits of the whole key are set by the feature. Specifically, in the case of single atom descriptors, nl is 0 and n2 and n3 encode one or two properties of the atom for atom pair descriptors, nl encodes the number of bonds (topological distance) between the atoms, while nl and n3 encode the property values of the two atoms finally, for custom atom environment descriptors, nl is equal to 7, while n2 encodes the specific atom environment and n3 encodes the property of the atom in the center of that environment. The number n4 encodes the number of occurrences in the molecule of the considered feature. The number n5 is used to specify the number of bits that are set, while n6 is a flag indicating whether or not hashing is allowed the final three numbers, n7, n8, and n9 identify the bits in the structural key. 

In the study of Bruggemann et al. (2001b) a series of synthesis specific descriptors, i.e. simple structural descriptors such as the number of specific atoms and the number of specific bonds were included in the analyses along with graph theoretical and quantum chemical descriptors. On this basis a 6-step procedure was developed to solve inverse QSAR problems. 

Stereochemistry is expressed in C A index names by three methods stereoparents, systematic stereodescriptors, and descriptors for co-ordination compounds. We have developed a separate procedure for converting each of these three types of stereodescriptors to atom/bond specific descriptors in the Registry connection table. Table 1 gives statistics for the Registry File as of February, 1990. At that time, the total file contained over 10 million registered substances. 

This finishes the specification of the atoms which constitute a compound. Again, the definition was simplified, and the specification of both atom symbol and atom attribute was not further elaborated. For atom symbol, the terms should be self-explaining. The atom attributes allow a very precise specification of the properties of an atom. The attributes atom stereo descriptor, coordination, and atom parity flag are part of the stereochemistry description. The different counts and the valency are just positive integers, the cartesian coordinates are provided for a graphical display of a structure with a suitable program. 

If one considers only the electron densities in the highest occupied and lowest unoccupied MOs, the so-called electrophilic and nucleophilic frontier orbital densities result. These descriptors suppose that the HOMO and LUMO are far more important than the other MOs in determining the position and likelihood of electrophilic or nucleophilic attack. Again, when used in the manner discussed previously, these atom-specific indices become whole molecule descriptors. 

In chemoinformatics, chirality is taken into account by many structural representation schemes, in order that a specific enantiomer can be imambiguously specified. A challenging task is the automatic detection of chirality in a molecular structure, which was solved for the case of chiral atoms, but not for chirality arising from other stereogenic units. Beyond labeling, quantitative descriptors of molecular chirahty are required for the prediction of chiral properties such as biological activity or enantioselectivity in chemical reactions) from the molecular structure. These descriptors, and how chemoinformatics can be used to automatically detect, specify, and represent molecular chirality, are described in more detail in Chapter 8. 

A limitation of the ap and tt descriptors is the specificity of the atom typing, e.g., benzoic acid and phenyltetrazole would not be perceived as very similar, even though carboxylates and tetrazoles are both anions at physiological pH. 

A common feature of the various methods that we have developed for the calculation of electronic effects in organic molecules is that they start from fundamental atomic data such as atomic ionization potentials and electron affinities, or atomic polarizability parameters. These atomic data are combined according to specific physical models, to calculate molecular descriptors which take account of the network of bonds. In other words, the constitution of a molecule (the topology) determines the way the procedures (algorithms) walk through the molecule. Again, as previously mentioned, the calculations are performed on the entire molecule. 

Chemical Information, Irvine CA Tripos, Inc. St. Louis MO), similarity searching can be carried out around a well-defined compound class using local descriptors such as atom pairs [46, 47] or topomeric shape [48, 49]. Also, ligand-based pharmacophore searches are able to identify follow-up compounds that are less obvious and more diverse than similarity searches [30, 50-54]. The problem with the latter methods is defining the molecular shape or pharmacophore specifically enough to be useful when there are few hits within a compound class and they cannot be reliably aligned (as is often the case for NMR hits in the absence of detailed structural information). 

---