Prediction of anticancer molecules using hybrid model developed on molecules screened against NCI-60 cancer cell lines

Abstract

BackgroundIn past, numerous quantitative structure-activity relationship (QSAR) based models have been developed for predicting anticancer activity for a specific class of molecules against different cancer drug targets. In contrast, limited attempt have been made to predict the anticancer activity of a diverse class of chemicals against a wide variety of cancer cell lines. In this study, we described a hybrid method developed on thousands of anticancer and non-anticancer molecules tested against National Cancer Institute (NCI) 60 cancer cell lines.ResultsOur analysis of anticancer molecules revealed that majority of anticancer molecules contains 18–24 carbon atoms and are dominated by functional groups like R2NH, R3N, ROH, RCOR, and ROR. It was also observed that certain substructures (e.g., 1-methoxy-4-methylbenzene, 1-methoxy benzene, Nitrobenzene, Indole, Propenyl benzene) are more abundant in anticancer molecules. Next, we developed anticancer molecule prediction models using various machine-learning techniques and achieved maximum matthews correlation coefficient (MCC) of 0.81 with 90.40 % accuracy using support vector machine (SVM) based models. In another approach, a novel similarity or potency score based method has been developed using selected fragments/fingerprints and achieved maximum MCC of 0.82 with 90.65 % accuracy. Finally, we combined the strength of above methods and developed a hybrid method with maximum MCC of 0.85 with 92.47 % accuracy.ConclusionsWe developed a hybrid method utilizing the best of machine learning and potency score based method. The highly accurate hybrid method can be used for classification of anticancer and non-anticancer molecules. In order to facilitate scientific community working in the field of anticancer drug discovery, we integrate hybrid and potency method in a web server CancerIN. This server provides various facilities that includes; virtual screening of anticancer molecules, analog based drug design, and similarity with known anticancer molecules (http://crdd.osdd.net/oscadd/cancerin).Electronic supplementary materialThe online version of this article (doi:10.1186/s12885-016-2082-y) contains supplementary material, which is available to authorized users.

Conformation‐Independent QSAR Study on Human Epidermal Growth Factor Receptor‐2 (HER2) Inhibitors

“…MACCSFP28 fingerprint accounts for a CH 2 moiety bonded to heteroatoms, for example nitrogen, whereas PubchemFP539 is indicator of a nitrogen double‐bonded to a carbon, as in an aromatic or tertiary amine. Recently secondary and tertiary amino groups have been reported as the most frequent moieties found in anticancer drugs tested against NCI‐60 cell lines in a QSAR study using a dataset of 8565 molecules . The molecular descriptors and fingerprints appearing in the model of Eq.…”

“…Recently secondary and tertiary amino groups have been reported as the most frequent moieties found in anticancer drugs tested against NCI-60 cell lines in a QSAR study using a dataset of 8565 molecules. [77] The molecular descriptors and fingerprints appearing in the model of Eq. 1 suggest that the inhibitory activity of HER2 is affected by the aromatic character of the compounds, its ionization potential and the presence of nitrogen atoms in the structure, particularly those forming amino groups and not associated directly to oxygen.…”

Design and Anticancer Activity Prediction of Dihyropyrimidinone Based Novel Inhibitors of P53-Mdm2 Interaction

“…The potency score provides the distance of query molecule with active and inactive anticancer inhibitors. Thus, a molecule having high potency score is more similar to active anticancer inhibitors as compare to inactive anticancer inhibitors [9]. In anticancer activity study compound, L9-L11 showed highest PS values (0.22) among the series as well as reference compounds (0.15-0.16).…”

“…The anticancer activity of each compound (including reference) was determined in comparison with most active anticancer agent and inactive anticancer agents in National Cancer Institute (NCI) database using "Cancer IN" server [9]. SMILE files of each compound were generated using Chem3D were used as input files for "Cancer IN."…”

“…Where Max is maximum score, H a T s1 indicates Tanimoto or Jaccard similarity score, H a T s0 is highest similarity score between the query and most similar anticancer agents, H n T s1 and H n T s0 indicates similarity scores between the query and most similar nonanticancer agents [9,22]. The potency score values vary from −1 to +1.…”