In Silico HCT116 Human Colon Cancer Cell-Based Models En Route to the Discovery of Lead-Like Anticancer Drugs

Abstract

To discover new inhibitors against the human colon carcinoma HCT116 cell line, two quantitative structure–activity relationship (QSAR) studies using molecular and nuclear magnetic resonance (NMR) descriptors were developed through exploration of machine learning techniques and using the value of half maximal inhibitory concentration (IC50). In the first approach, A, regression models were developed using a total of 7339 molecules that were extracted from the ChEMBL and ZINC databases and recent literature. The performance of the regression models was successfully evaluated by internal and external validations, the best model achieved R2 of 0.75 and 0.73 and root mean square error (RMSE) of 0.66 and 0.69 for the training and test sets, respectively. With the inherent time-consuming efforts of working with natural products (NPs), we conceived a new NP drug hit discovery strategy that consists in frontloading samples with 1D NMR descriptors to predict compounds with anticancer activity prior to bioactivity screening for NPs discovery, approach B. The NMR QSAR classification models were built using 1D NMR data (1H and 13C) as descriptors, from 50 crude extracts, 55 fractions and five pure compounds obtained from actinobacteria isolated from marine sediments collected off the Madeira Archipelago. The overall predictability accuracies of the best model exceeded 63% for both training and test sets.

A Computer-Driven Approach to Discover Natural Product Leads for Methicillin-Resistant Staphylococcus aureus Infection Therapy

“…Following our previous work that modeling the anticancer activity against HCT116 [37], the current results suggest as well that the chemoinformatics QSAR approach relying on a ligand-based methodology either based on the molecular structures or the NMR spectra, corroborated with an experimental approach, could be used to predict new inhibitory compounds against MRSA. To our knowledge, the QSAR regression model developed here, approach A, is the largest study ever performed with regard both to the number of compounds involved and to the number of structural families involved in the modeling of the antibacterial activity against MRSA [19,20,25].…”

“…This result is not surprising since the genus Streptomyces over the last decades has stirred huge interest as a source of bioactive compounds, more than 60% of all known antibiotics have been isolated from streptomycetes [39], and moreover a similarly result was obtained by us for the anticancer screening against HCT116 in a previously study from actinobacteria isolated from marine sediments collected off the Madeira Archipelago [37].…”

“…To our knowledge, the QSAR regression model developed here, approach A, is the largest study ever performed with regard both to the number of compounds involved and to the number of structural families involved in the modeling of the antibacterial activity against MRSA [19,20,25]. The NMR QSAR classification model, approach B, was extended to a high number of samples containing additional 45 pure compounds and therefore the overall predictability accuracies (Q) were improved as compared with those obtained in our previously work [37]. The Q of the best model exceeded 77% and 63% for test set in MRSA and HCT116 modeling, respectively.…”

“…These models were developed using in total 6645 molecules that were extracted from the ChEMBL, PubChem, and ZINC databases and recent literature indexed in Web of Science. In order to reduce the time consumption and biological activity screening-associated costs that are associated to NP discovery, a new NP drug discovery strategy similarly to that we had recently developed for anticancer activity [37] was developed. In the approach B was used NMR descriptors for building NMR QSAR models to predict compounds with antibacterial activity prior to bioactivity screening for NP discovery.…”

Exploiting machine learning for end-to-end drug discovery and development

“…In summary, while much of the focus of this review has been on cheminformatics, we propose that many areas across the pharmaceutical R&D spectrum and outside of it are ripe for machine learning (Table 1). Machine learning can learn from almost any data type, such as that from research papers, patient records, images, genes, symptoms, diseases, proteins, tissues, species and drug candidates or compounds that have been shown to affect any of the preceding 74 . We could also imagine a complex interaction network between proteins upstream and downstream in the pathway that might dictate if the drug/s will work.…”