Predicting mTOR Inhibitors with a Classifier Using Recursive Partitioning and Naïve Bayesian Approaches

Abstract

BackgroundMammalian target of rapamycin (mTOR) is a central controller of cell growth, proliferation, metabolism, and angiogenesis. Thus, there is a great deal of interest in developing clinical drugs based on mTOR. In this paper, in silico models based on multi-scaffolds were developed to predict mTOR inhibitors or non-inhibitors.MethodsFirst 1,264 diverse compounds were collected and categorized as mTOR inhibitors and non-inhibitors. Two methods, recursive partitioning (RP) and naïve Bayesian (NB), were used to build combinatorial classification models of mTOR inhibitors versus non-inhibitors using physicochemical descriptors, fingerprints, and atom center fragments (ACFs).ResultsA total of 253 models were constructed and the overall predictive accuracies of the best models were more than 90% for both the training set of 964 and the external test set of 300 diverse compounds. The scaffold hopping abilities of the best models were successfully evaluated through predicting 37 new recently published mTOR inhibitors. Compared with the best RP and Bayesian models, the classifier based on ACFs and Bayesian shows comparable or slightly better in performance and scaffold hopping abilities. A web server was developed based on the ACFs and Bayesian method (http://rcdd.sysu.edu.cn/mtor/). This web server can be used to predict whether a compound is an mTOR inhibitor or non-inhibitor online.Conclusion In silico models were constructed to predict mTOR inhibitors using recursive partitioning and naïve Bayesian methods, and a web server (mTOR Predictor) was also developed based on the best model results. Compound prediction or virtual screening can be carried out through our web server. Moreover, the favorable and unfavorable fragments for mTOR inhibitors obtained from Bayesian classifiers will be helpful for lead optimization or the design of new mTOR inhibitors.

A de novo substructure generation algorithm for identifying the privileged chemical fragments of liver X receptorβ agonists

“…Over the past decades, many substructure generation approaches have been reported, such as empirical search keys 32 , algorithm-based atom center fragments 13 , 33 , 34 , fingerprints ( http://www.daylight.com/ ) 15 , 35 . De novo substructures are derived by algorithms with a given minimal support threshold (popularity threshold).…”

Discovering new mTOR inhibitors for cancer treatment through virtual screening methods and in vitro assays

“…2 . To evaluate the novelty of these hits with respect to known mTOR kinase inhibitors, pairwise Tanimoto similarity indices between these hits and mTOR inhibitors obtained from ChEMBL (IC 50 < 10 μM, Figure S3 ) 18 were calculated based on the FCFP_4 fingerprint via the “Find Similar Molecules by Fingerprints protocol” in Discovery Studio 3.5 (Accelrys Inc., San Diego, USA). As shown in Figure S3 , these hits have low Tanimoto similarities (0.13 ~ 0.38, except 25 of 0.421) with the known mTOR inhibitors.…”

“…In the previous work, we developed an in silico method to predict mTOR inhibitors with multiple classification approaches including recursive partitioning (RP), naïve Bayesian (NB) learning 18 using Atom Center Fragments (ACFs) as the features. The method has been validated for being capable of hopping new mTOR inhibitor scaffolds 18 . In this study, we continued our earlier efforts aimed at identifying and characterizing novel mTOR inhibitors.…”

Overview of Methods and Strategies for Conducting Virtual Small Molecule Screening

“…A NB classifier needs to be trained with sets of known active and inactive molecules. NB is computationally fast, and very efficient at finding new actives (Wang et al, 2014). Compared to a simple similarity search, a NB classifier is more flexible and is able to find molecules that are somewhat different from the known actives.…”