Search

Traditional small-molecule drug discovery is a time-consuming and costly endeavor. High-throughput chemical screening can only assess a tiny fraction of drug-like chemical space. The strong predictive power of modern machine-learning methods for virtual chemical screening enables training models on known active and inactive compounds and extrapolating to much larger chemical libraries. However, there has been limited experimental validation of these methods in practical applications on large commercially available or synthesize-on-demand chemical libraries. Through a prospective evaluation with the bacterial protein-protein interaction PriA-SSB, we demonstrate that ligand-based virtual screening can identify many active compounds in large commercial libraries. We use cross-validation to compare different types of supervised learning models and select a random forest (RF) classifier as the best model for this target. When predicting the activity of more than 8 million compounds from Aldrich Market Select, the RF substantially outperforms a naïve baseline based on chemical structure similarity. 48% of the RF's 701 selected compounds are active. The RF model easily scales to score one billion compounds from the synthesize-on-demand Enamine REAL database. We tested 68 chemically diverse top predictions from Enamine REAL and observed 31 hits (46%), including one with an IC 50 value of 1.3 μM.

Virtual (computational) high-throughput screening provides a strategy for prioritizing compounds for experimental screens, but the choice of virtual screening algorithm depends on the data set and evaluation strategy. We consider a wide range of ligand-based machine learning and docking-based approaches for virtual screening on two protein-protein interactions, PriA-SSB and RMI-FANCM, and present a strategy for choosing which algorithm is best for prospective compound prioritization. Our workflow identifies a random forest as the best algorithm for these targets over more sophisticated neural network-based models. The top 250 predictions from our selected random forest recover 37 of the 54 active compounds from a library of 22,434 new molecules assayed on PriA-SSB. We show that virtual screening methods that perform well on public data sets and synthetic benchmarks, like multi-task neural networks, may not always translate to prospective screening performance on a specific assay of interest.

Screening of a marine natural products library for inhibitors of TGF-β revealed five pyrimidinedione derivatives, biemamides A-E (1-5). The structures were determined by 2D NMR and HRMS experiments; absolute configurations were established by advanced Marfey's analysis and ECD calculations. Biemamides A-E specifically inhibited in vitro TGF-β induced epithelial to mesenchymal transition in NMuMG cells. Additionally, using Caenorhabditis elegans, selected biemmamides were found to influence in vivo developmental processes related to body size regulation in a dose-dependent manner.

In structure-based virtual screening, compound ranking through a consensus of scores from a variety of docking programs or scoring functions, rather than ranking by scores from a single program, provides better predictive performance and reduces target performance variability. Here we compare traditional consensus scoring methods with a novel, unsupervised gradient boosting approach. We also observed increased score variation among active ligands and developed a statistical mixture model consensus score based on combining score means and variances. To evaluate performance, we used the common performance metrics ROCAUC and EF1 on 21 benchmark targets from DUD-E. Traditional consensus methods, such as taking the mean of quantile normalized docking scores, outperformed individual docking methods and are more robust to target variation. The mixture model and gradient boosting provided further improvements over the traditional consensus methods. These methods are readily applicable to new targets in academic research and overcome the potentially poor performance of using a single docking method on a new target.

The biosynthetic gene cluster for tautomycetin (TTN), a highly potent and selective protein phosphatase (PP) inhibitor isolated from Streptomyces griseochromogenes, has recently been cloned and sequenced. To better understand the transformations responsible for converting the post-polyketide synthase product into the exciting anticancer and immunosuppressive chemotherapeutic candidate TTN, we produced and characterized new analogues resulting from inactivation of two genes, ttnD and ttnF, in S. griseochromogenes. Inactivation of ttnD and ttnF, which encode for putative decarboxylase and dehydratase enzymes, respectively, afforded mutant strains SB13013 and SB13014. The DeltattnD mutant SB13013 accumulated four new TTN analogues, TTN D-1, TTN D-2, TTN D-3, and TTN D-4, whereas the DeltattnF mutant accumulated only one new TTN analogue, TTN F-1. The accumulation of these new TTN analogues defines the function of TtnD and TtnF and the timing of their chemistries in relation to installation of the C5 ketone moiety within TTN. Notably, all new analogues possess a structurally distinguishing carboxylic acid moiety, revealing that TtnD apparently cannot catalyze decarboxylation in the absence of TtnF. Additionally, cytotoxicity and PP inhibition assays reveal the importance of the functional groups installed by TtnDF and, consistent with earlier proposals, the C2''-C5 fragment of TTN to be a critical structural determinant behind the important and unique PP-1 selectivity displayed by TTN.

The biosynthetic gene cluster for tautomycin (TTM), a potent protein phosphatase (PP) inhibitor has recently been characterized. Inactivation of ttmM, which encodes a putative C3' hydroxylase, afforded mutant SB6005 which accumulated three new 3'-deshydroxy TTM analogs, supporting the function of TtmM and the previously proposed linear pathway for TTM biosynthesis. Bioassays reveal the importance of the C3' OH moiety in PP inhibition and that PP inhibition is not the exclusive mechanism driving TTM-induced cell death.

Migrastatin (1), iso-migrastatin (5) and lactimidomycin (7) are all glutarimide-containing polyketides known for their unique structures and cytotoxic activities against human cancer cell lines. Migrastatin, a strong inhibitor of tumor cell migration, has been an important lead in the development of antimetastatic agents. Yet studies of the related 12-membered macrolides iso-migrastatin, lactimidomycin, and related analogues have been hampered by their limited availability. We report here the production, isolation, structural characterization, and biological activities of iso-migrastatin, lactimidomycin, and 23 related congeners. Our studies showed that, as a family, the glutarimide-containing 12-membered macrolides are extremely potent cell migration inhibitors with some members displaying activity on par or superior to that of migrastatin as exemplified by compounds 5, 7, and 9-12. On the basis of these findings, the structures and activity of this family of compounds as cell migration inhibitors are discussed.

Using STM, infrared absorption reflection spectroscopy experiments and density functional calculations we show that low temperature adsorption of CO on gold surfaces modified by vacancy islands leads to morphological changes and the formation of nanosized Au particles. These results demonstrate a dynamic response of a surface during adsorption with consequences for the surface reactivity.

The reaction of 70 unprotected, diversely functionalized free reducing sugars with methoxyamine-appended colchicine led to the production of a 58-member glycorandomized library. High-throughput cytotoxicity assays revealed glycosylation to modulate specificity and potency. Library members were also identified which, unlike the parent natural product (a destabilizer), stabilized in vitro tubulin polymerization in a manner similar to taxol. This study highlights a simple extension of neoglycorandomization toward amine-bearing scaffolds and the potential benefit of glycosylating nonglycosylated natural products.