Application of supervised Kohonen map and counter propagation neural network for classification of nucleic acid structures based on their circular dichroism spectra

Deriving general structure-activity/selectivity relationship patterns for different subfamilies of cyclin-dependent kinase inhibitors using machine learning methods

Cyclin-dependent kinases (CDKs) play essential roles in regulating the cell cycle and are among the most critical targets for cancer therapy and drug discovery. The primary objective of this research is to derive general structure-activity relationship (SAR) patterns for modeling the selectivity and activity levels of CDK inhibitors using machine learning methods. To accomplish this, 8592 small molecules with different binding affinities to CDK1, CDK2, CDK4, CDK5, and CDK9 were collected from Binding DB, and a diverse set of descriptors was calculated for each molecule. The supervised Kohonen networks (SKN) and counter propagation artificial neural networks (CPANN) models were trained to predict the activity levels and therapeutic targets of the molecules. The validity of models was confirmed through tenfold cross-validation and external test sets. Using selected sets of molecular descriptors (e.g. hydrophilicity and total polar surface area) we derived activity and selectivity maps to elucidate local regions in chemical space for active and selective CDK inhibitors. The SKN models exhibited prediction accuracies ranging from 0.75 to 0.94 for the external test sets. The developed multivariate classifiers were used for ligand-based virtual screening of 2 million random molecules of the PubChem database, yielding areas under the receiver operating characteristic curves ranging from 0.72 to 1.00 for the SKN model. Considering the persistent challenge of achieving CDK selectivity, this research significantly contributes to addressing the issue and underscores the paramount importance of developing drugs with minimized side effects.

CD-NuSS: A Web Server for the Automated Secondary Structural Characterization of the Nucleic Acids from Circular Dichroism Spectra Using Extreme Gradient Boosting Decision-Tree, Neural Network and Kohonen Algorithms

Nucleic acids exhibit a repertoire of conformational preference depending on the sequence and environment. Circular dichroism (CD) is an essential and valuable tool for monitoring such secondary structural conformations of nucleic acids. Nonetheless, the CD spectral diversity associated with these structures poses a challenge in obtaining the quantitative information about the secondary structural content of a given CD spectrum. To this end, the competence of the extreme gradient boosting decision-tree (XGBoost), Kohonen and neural network (nnet) algorithms have been exploited here to predict the diverse secondary structures of nucleic acids. A curated library of 450 CD spectra corresponding to 16 different secondary structures of nucleic acids has been created and used as a training dataset. The hyper-parameters corresponding to the aforementioned algorithms have been optimized using holdout and k-fold (here, k = 5) cross-validation methods. For a test dataset of 150 CD spectra, both the nnet and XGBoost algorithms have exhibited nearly similar prediction accuracy in the range of 85% and 87% (the latter exhibited a slightly higher prediction accuracy). Thus, the nnet and XGBoost algorithms tested here can be employed for predicting the hybrid nucleic acid topologies in future. For the sake of accessibility, the entire process has been automated and implemented as a webserver, called CD-NuSS (CD to nucleic acids secondary structure) and is freely accessible at https://project.iith.ac.in/cdnuss/.