Pharmacophore fingerprinting. 2. Application to primary library design.

A methodology for pharmacophore fingerprinting (PharmPrint), previously described in the context of QSAR, has been used to address the issues involved in primary library design. A subset of the MDDR (MDDR9104) has been used to define a reference set of bioactive molecules. A statistic has been devised to measure the discriminating power of molecular descriptors using the target class assignments for this set, for which the PharmPrint fingerprint outperformed other descriptors. A principal components analysis (PCA) of the fingerprints for the MDDR9104 produces a low dimensional representation within which molecular properties and other libraries can be visualized and explored. PCA calculations on subsets of classes show that this space is robust to the addition of new classes, suggesting that pharmacophoric space is finite and rapidly converging. We demonstrate the application of the PharmPrint methodology to the analysis and design of virtual combinatorial libraries using common scaffolds and building blocks.

Pharmacophore fingerprinting. 1. Application to QSAR and focused library design.

A new method of rapid pharmacophore fingerprinting (PharmPrint method) has been developed. A basis set of 10,549 three-point pharmacophores has been constructed by enumerating several distance ranges and pharmacophoric features. Software has been developed to assign pharmacophoric types to atoms in chemical structures, generate multiple conformations, and construct the binary fingerprint according to the pharmacophores that result. The fingerprint is used as a descriptor for developing a quantitative structure-activity relationship (QSAR) model using partial least squares. An example is given using sets of ligands for the estrogen receptor (ER). The result is compared with previously published results on the same data to show the superiority of a full 3D, conformationally flexible approach. The QSAR model can be readily interpreted in structural/chemical terms. Further examples are given using binary activity data and some of our novel in-house compounds, which show the value of the model when crossing compound classes.

Prediction of human intestinal absorption of drug compounds from molecular structure.

The absorption of a drug compound through the human intestinal cell lining is an important property for potential drug candidates. Measuring this property, however, can be costly and time-consuming. The use of quantitative structure-property relationships (QSPRs) to estimate percent human intestinal absorption (%HIA) is an attractive alternative to experimental measurements. A data set of 86 drug and drug-like compounds with measured values of %HIA taken from the literature was used to develop and test a QSPR mode. The compounds were encoded with calculated molecular structure descriptors. A nonlinear computational neural network model was developed by using the genetic algorithm with a neural network fitness evaluator. The calculated %HIA (cHIA) model performs wells, with root-mean-square (rms) errors of 9.4%HIA units for the training set, 19.7%HIA units for the cross-validation (CV) set, and 16.0%HIA units for the external prediction set.

Predicting protein secondary structure content. A tandem neural network approach.

A priori knowledge of secondary structure content can be of great use in theoretical and experimental determination of protein structure. We present a method that uses two computer-simulated neural networks placed in "tandem" to predict the secondary structure content of water-soluble, globular proteins. The first of the two networks, NET1, predicts a protein's helix and strand content given information about the protein's amino acid composition, molecular weight and heme presence. Because NET1 contained more adjustable parameters (network weights) than learning examples, this network experienced problems with memorization, which is the inability to generalize onto new, never-seen-before examples. To overcome this problem, we designed a second network, NET2, which learned to determine when NET1 was in a state of generalization. Together, these two networks produce prediction errors as low as 5.0% and 5.6% for helix and strand content, respectively, on a set of protein crystal structures bearing little homology to those used in network training. A comparison between three other methods including a multiple linear regression analysis, a non-hidden-node network analysis and a secondary structure assignment analysis reveals that our tandem neural network scheme is, indeed, the best method for predicting secondary structure content. The results of our analysis suggest that the knowledge of sequence information is not necessary for highly accurate predictions of protein secondary structure content.

Predicting surface exposure of amino acids from protein sequence.

The amino acid residues on a protein surface play a key role in interaction with other molecules, determined many physical properties, and constrain the structure of the folded protein. A database of monomeric protein crystal structures was used to teach computer-simulated neural networks rules for predicting surface exposure from local sequence. These trained networks are able to correctly predict surface exposure for 72% of residues in a testing set using a binary model, (buried/exposed) and for 54% of residues using a ternary model (buried/intermediate/exposed). In the ternary model, only 11% of the exposed residues are predicted as buried and only 5% of the buried residues are predicted as exposed. Also, since the networks are able to predict exposure with a quantitative confidence estimate, it is possible to assign exposure for over half of the residues in a binary model with greater than 80% accuracy. Even more accurate predictions are obtained by making a consensus prediction of exposure for a homologous family. The effect of the local environment of an amino acid on its accessibility, though smaller than expected, is significant and accounts for the higher success rate of prediction than obtained with previously used criteria. In the absence of a three-dimensional structure, the ability to predict surface accessibility of amino acids directly from the sequence is a valuable tool in choosing sites of chemical modification or specific mutations and in studies of molecular interaction.

Prediction of the disulfide-bonding state of cysteine in proteins.

The bonding states of cysteine play important functional and structural roles in proteins. In particular, disulfide bond formation is one of the most important factors influencing the three-dimensional fold of proteins. Proteins of known structure were used to teach computer-simulated neural networks rules for predicting the disulfide-bonding state of a cysteine given only its flanking amino acid sequence. Resulting networks make accurate predictions on sequences different from those used in training, suggesting that local sequence greatly influences cysteines in disulfide bond formation. The average prediction rate after seven independent network experiments is 81.4% for disulfide-bonded and 80.0% for non-disulfide-bonded scenarios. Predictive accuracy is related to the strength of network output activities. Network weights reveal interesting position-dependent amino acid preferences and provide a physical basis for understanding the correlation between the flanking sequence and a cysteine's disulfide-bonding state. Network predictions may be used to increase or decrease the stability of existing disulfide bonds or to aid the search for potential sites to introduce new disulfide bonds.