No abstract

Computational methods have been widely used in drug design. The recent developments in machine learning techniques and the ever-growing chemical and biological databases are fertile ground for discoveries in this area. In this study, we evaluated the performance of Deep Learning models in comparison to Random Forest, and Support Vector Regression for predicting the biological activity (pIC50) of ALK-5 inhibitors as candidates to treat cancer. The generalization power of the models was assessed by internal and external validation procedures. A deep neural network model obtained the best performance in this comparative study, achieving a coefficient of determination of 0.658 on the external validation set with mean square error and mean absolute error of 0.373 and 0.450, respectively. Additionally, the relevance of the chemical descriptors for the prediction of biological activity was estimated using Permutation Importance. We can conclude that the forecast model obtained by the deep neural network is suitable for the problem and can be employed to predict the biological activity of new ALK-5 inhibitors.

A key step in the biosynthesis of numerous polyketides is the stereospecific formation of a spiroacetal (spiroketal). We report here that spiroacetal formation in the biosynthesis of the macrocyclic polyketides ossamycin and oligomycin involves catalysis by a novel spiroacetal cyclase. OssO from the ossamycin biosynthetic gene cluster (BGC) is homologous to OlmO, the product of an unannotated gene from the oligomycin BGC. The deletion of olmO abolished oligomycin production and led to the isolation of oligomycin-like metabolites lacking the spiroacetal structure. Purified OlmO catalyzed complete conversion of the major metabolite into oligomycin C. Crystal structures of OssO and OlmO reveal an unusual 10-strand β-barrel. Three conserved polar residues are clustered together in the β-barrel cavity, and site-specific mutation of any of these residues either abolished or substantially diminished OlmO activity, supporting a role for general acid/general base catalysis in spiroacetal formation.

HER-2 and EGFR are biological targets related to the development of cancer and the discovery and/or development of a dual inhibitor could be a good strategy to design an effective drug candidate. In this study, analyses of the chemical properties of a group of substances having affinity for both HER-2 and EGFR were carried out with the aim of understanding the main factors involved in the interaction between these inhibitors and the biological targets. Comparative analysis of molecular interaction fields (CoMFA) and comparative molecular similarity index analysis (CoMSIA) techniques were applied on 63 compounds. From CoMFA analyses, we found for both HER-2 (r2 calibration = 0.98 and q2cv = 0.83) and EGFR (r2 calibration = 0.98 and q2cv = 0.73) good predictive models. Good models for CoMSIA technique have also been found for HER-2 (r2 calibration = 0.92 and q2cv = 0.74) and EGFR (r2 calibration = 0.97 and q2cv = 0.72). The constructed models could indicate some important characteristics for the inhibition of the biological targets. New compounds were proposed as candidates to inhibit both proteins. Therefore, this study may guide future projects for the development of new drug candidates for the treatment of breast cancer.

No abstract