A method for automatic generation of novel chemical structures and its potential applications to drug discovery

A new formalism for molecular shape description is described. The formalism, based on considering each molecule as a collection of its 3-atom submolecules, is applied to both the graph theory and geometrical coordinate representations of molecules. The timing results for shape description of several databases indicate that this new method is applicable to large databases. Furthermore, results from clustering a small database show good agreement with clustering results obtained by a distance-matching algorithm.

Section: Resultsmentioning

confidence: 99%

A fast and efficient method for 2D and 3D molecular shape description

Bemis

Kuntz

1992

“…5. T[1]-T [5] have many positive atom pairs in common. We interpret this as follows: compounds that have two or more of the same descriptors simultaneously are more likely to be active than expected from the trend vector prediction.…”

Section: Samplsmentioning

confidence: 99%

“…(The CASE system [3] has a similar philosophy.) The trend vector, used with the descriptors described above, has proved to be useful in a number of applications [4][5][6][7]. One application is the evaluation of chemical structures as candidates for synthesis.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Extending the trend vector: The trend matrix and sample-based partial least squares

Sheridan

Nachbar

Bush

1994

Trend vector analysis [Carhart, R.E. et al., J. Chem. Inf. Comput. Sci., 25 (1985) 64], in combination with topological descriptors such as atom pairs, has proved useful in drug discovery for ranking large collections of chemical compounds in order of predicted biological activity. The compounds with the highest predicted activities, upon being tested, often show a several-fold increase in the fraction of active compounds relative to a randomly selected set. A trend vector is simply the one-dimensional array of correlations between the biological activity of interest and a set of properties or 'descriptors' of compounds in a training set. This paper examines two methods for generalizing the trend vector to improve the predicted rank order. The trend matrix method finds the correlations between the residuals and the simultaneous occurrence of descriptors, which are stored in a two-dimensional analog of the trend vector. The SAMPLS method derives a linear model by partial least squares (PLS), using the 'sample-based' formulation of PLS [Bush, B.L. and Nachbar, R.B., J. Comput.-Aided Mol. Design, 7 (1993) 587] for efficiency in treating the large number of descriptors. PLS accumulates a predictive model as a sum of linear components. Expressed as a vector of prediction coefficients on properties, the first PLS component is proportional to the trend vector. Subsequent components adjust the model toward full least squares. For both methods the residuals decrease, while the risk of overfitting the training set increases. We therefore also describe statistical checks to prevent overfitting. These methods are applied to two data sets, a small homologous series of disubstituted piperidines, tested on the dopamine receptor, and a large set of diverse chemical structures, some of which are active at the muscarinic receptor. Each data set is split into a training set and a test set, and the activities in the test set are predicted from a fit on the training set. Both the trend matrix and the SAMPLS approach improve the predictions over the simple trend vector. The SAMPLS approach is superior to the trend matrix in that it requires much less storage and CPU time. It also provides a useful set of axes for visualizing properties of the compounds. We describe a randomization method to determine the optimum number of PLS components that is very much faster for large training sets than leave-one-out cross-validation.

“…Programs like GENSTAR [20] should therefore be seen as (to quote the authors) 'molecular sketch pads'; a rich source of creative ideas for a medicinal chemist to explore. A similar procedure for building structures in two dimensions (with connectivities but no coordinates) has been described by Nilakantan et al [22].…”

Section: Structure Generation By Fragment Joiningmentioning

confidence: 99%

Current methods for site-directed structure generation

Lewis¹,

Leach

1994

There has been a rapid growth of interest in techniques for site-directed drug design, fueled by the increasing availability of structural models of proteins of therapeutic importance, and by studies reported in the literature showing that potent chemical leads can be obtained by these techniques. Structure generation programs offer the prospect of discovering highly original lead structures from novel chemical families. Due to the fact that this technique is more-or-less still in its infancy, there are no case studies available that demonstrate the use of structure generation programs for site-directed drug design. Such programs were first proposed in 1986, and became commercially available in early 1992. They have shown their ability to reproduce, or suggest reasonable alternatives for, ligands in well-defined binding sites. This brief review will discuss the recent advances that have been made in the field of site-directed structure generation.