A lazy learning-based QSAR classification study for screening potential histone deacetylase 8 (HDAC8) inhibitors

Abstract

Histone deacetylases 8 (HDAC8) is an enzyme repressing the transcription of various genes including tumour suppressor gene and has already become a target of human cancer treatment. In an effort to facilitate the discovery of HDAC8 inhibitors, two quantitative structure-activity relationship (QSAR) classification models were developed using K nearest neighbours (KNN) and neighbourhood classifier (NEC). Molecular descriptors were calculated for the data set and database compounds using ADRIANA.Code of Molecular Networks. Principal components analysis (PCA) was used to select the descriptors. The developed models were validated by leave-one-out cross validation (LOO CV). The performances of the developed models were evaluated with an external test set. Highly predictive models were used for database virtual screening. Furthermore, hit compounds were subsequently subject to molecular docking. Five hits were obtained based on consensus scoring function and binding affinity as potential HDAC8 inhibitors. Finally, HDAC8 structures in complex with five hits were also subjected to 5 ns molecular dynamics (MD) simulations to evaluate the complex structure stability. To the best of our knowledge, the NEC classification model used in this study is the first application of NEC to virtual screening for drug discovery.

High‐dimensional QSAR classification model for anti‐hepatitis C virus activity of thiourea derivatives based on the sparse logistic regression model with a bridge penalty

“…In terms of the MCC, on the other hand, the obtained value was considerably higher for the SLRM‐Bridge in the training datasets compared with those for the SLRM‐SCAD and SLRM‐ALASSO. Generally, a method with a higher MCC value is considered to be a more predictive classification method . Moreover, the constructed QSAR classification model using the SLRM‐Bridge also classified the test dataset with a very low misclassification rate, showing a MISS of 2.73% compared with 16.66% in the SLRM‐SCAD and 19.44% in the SLRM‐ALASSO.…”

Comparative assessment of favipiravir and remdesivir against human coronavirus NL63 in molecular docking and cell culture models

“…Based on the formed conformations, we can clearly visualize the binding of the two drugs to the RdRp of the two coronaviruses. LibDock score is a hallmark of binding affinity: the greater the LibDock score, the better the binding affinity 29 . Our simulation indicated a similar affinity of favipiravir in binding to the SARS-CoV-2 and HCoV-NL63 RdRp with LibDock scores of 75 and 74, respectively.…”

QSAR Models for Active Substances against Pseudomonas aeruginosa Using Disk-Diffusion Test Data

“…In terms of balanced accuracy, among individual models, the best performance was provided by KNN, logistic regression and decision tree classifier, but the ensemble method had superior results to individual models in this respect (the only model with BA generally over 70% in nested cross-validation, as well as on the external test set). KNN models have been successfully applied for other QSAR models, e.g., for different histone deacetylase inhibitors [38,39] or to predict binding affinity for different G-Protein Coupled Receptors (GPCRs) [40]. Logistic regression with regularization, although a relatively simple algorithm, has been shown to have similarly good performance as more sophisticated algorithms in QSAR models [41].…”

“…generally over 70% in nested cross-validation, as well as on the external test set). KNN models have been successfully applied for other QSAR models, e.g., for different histone deacetylase inhibitors [38,39] or to predict binding affinity for different G-Protein Coupled Receptors (GPCRs) [40]. Logistic regression with regularization, although a relatively simple algorithm, has been shown to have similarly good performance as more sophisticated algorithms in QSAR models [41].…”