The dimethyl sulfoxide (DMSO) solubility data from Enamine and two UCB pharma compound collections were analyzed using 8 different machine learning methods and 12 descriptor sets. The analyzed data sets were highly imbalanced with 1.7–5.8% nonsoluble compounds. The libraries’ enrichment by soluble molecules from the set of 10% of the most reliable predictions was used to compare prediction performances of the methods. The highest accuracies were calculated using a C4.5 decision classification tree, random forest, and associative neural networks. The performances of the methods developed were estimated on individual data sets and their combinations. The developed models provided on average a 2-fold decrease of the number of nonsoluble compounds amid all compounds predicted as soluble in DMSO. However, a 4–9-fold enrichment was observed if only 10% of the most reliable predictions were considered. The structural features influencing compounds to be soluble or nonsoluble in DMSO were also determined. The best models developed with the publicly available Enamine data set are freely available online at .
The conformations of the biologically active taxol analogs Taxotereo, 3R, 4R, and 4S, and the biologically inactive analog 3 S were evaluated in CDC13 and DMSO-water solution using 'H NMR coupling constant and NOESY data and molecular modeling. The solution structures of Taxoterem were very similar to those detected previously for taxol. The A-ring side chain conformations of analogs 3 and 4 could not be defined with the same precision as had been possible for taxol, but the conformational possibilities could be significantly limited by the data. Analogs 3R, 4R, and 4S (but not 3 s ) can mimic the dominant conforrnation of taxol in chloroform, but no logical relationship between biological activity and aqueous solution conformation could be detected. Chem. 72, 252 (1994).Faisant appel aux constantes de couplage en RMN du 'H obtenues pour des solutions de CDC13 ou aqueuses de DMSO, a des donnCes NEOSY et 3 de la modClisation molCculaire, on a CvaluC les conformations des analogues biologiquement actifs du taxol, les ~a x o t e r e @ 3 R , 4R et 4S, et de l'analogue biologiquement inactif 3 s . Les structures du ~a x o t e r e~ en solution sont trks semblables a celles dktectkes antkrieurement pour le taxol. Les conformations de la chaine laterale du cycle A des analogues 3 et 4 n'a pas CtC dkterminCe avec la prkcision qui avait CtC possible pour le taxol; toutefois, les possibilitCs conformationnelles ont pu Ctre skrieusement IimitCes par les donnCes. Les analogues 3R, 4R et 4S (mais pas le 3 s ) peuvent reproduire la conformation dominante du taxol dans le chloroforme; toutefois, on ne peut dCtecter aucune relation logique entre I'activitC biologique et la conformation en solution aqueuse.[Traduit par la rkdaction] Introduction The diterpenoid antineoplastic drug taxol (1, 2) is distinct from other anti-mitotic agents, such as the vinca alkaloids, in that it promotes the assembly and inhibits the disassembly of microtubules, both in vitro and in vivo (3). Whereas the structure-activity profile that is emerging3 hints at an intriguing interaction beiween taxol and its microtubular binding site(s).
Polylactide (PLA) was blended by conventional and reactive extrusion with limonene (LM) or myrcene (My) as bio-based plasticizers. As-processed blends were carefully analyzed by a multiscale and multidisciplinary approach to tentatively determine their chemical structure, microstructure, thermal properties, tensile and impact behaviors, and hydrothermal stability. The main results indicated that LM and My were efficient plasticizers for PLA, since compared to neat PLA, the glass transition temperature was reduced, the ultimate tensile strain was increased, and the impact strength was increased, independently of the type of extrusion. The addition of a free radical initiator during the extrusion of PLA/LM was beneficial for the mechanical properties. Indeed, the probable formation of local branched/crosslinked regions in the PLA matrix enhanced the matrix crystallinity, the tensile yield stress, and the tensile ultimate stress compared to the non-reactive blend PLA/LM, while the other properties were retained. For PLA/My blends, reactive extrusion was detrimental for the mechanical properties since My polymerization was accelerated resulting in a drop of the tensile ultimate strain and impact strength, and an increase of the glass transition temperature. Indeed, large inclusions of polymerized My were formed, decreasing the available content of My for the plasticization and enhancing cavitation from inclusion-matrix debonding.
BackgroundWe present “Ask Ernö”, a self-learning system for the automatic analysis of NMR spectra, consisting of integrated chemical shift assignment and prediction tools. The output of the automatic assignment component initializes and improves a database of assigned protons that is used by the chemical shift predictor. In turn, the predictions provided by the latter facilitate improvement of the assignment process. Iteration on these steps allows Ask Ernö to improve its ability to assign and predict spectra without any prior knowledge or assistance from human experts.ResultsThis concept was tested by training such a system with a dataset of 2341 molecules and their 1H-NMR spectra, and evaluating the accuracy of chemical shift predictions on a test set of 298 partially assigned molecules (2007 assigned protons). After 10 iterations, Ask Ernö was able to decrease its prediction error by 17 %, reaching an average error of 0.265 ppm. Over 60 % of the test chemical shifts were predicted within 0.2 ppm, while only 5 % still presented a prediction error of more than 1 ppm.ConclusionsAsk Ernö introduces an innovative approach to automatic NMR analysis that constantly learns and improves when provided with new data. Furthermore, it completely avoids the need for manually assigned spectra. This system has the potential to be turned into a fully autonomous tool able to compete with the best alternatives currently available.Graphical abstractSelf-learning loop. Any progress in the prediction (forward problem) will improve the assignment ability (reverse problem) and vice versa.Electronic supplementary materialThe online version of this article (doi:10.1186/s13321-016-0134-6) contains supplementary material, which is available to authorized users.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.