From UK-2A to florylpicoxamid: Active learning to identify a mimic of a macrocyclic natural product.

Scaffold replacement as part of an optimization process that requires maintenance of potency, desirable biodistribution, metabolic stability, and considerations of synthesis at very large scale is a complex challenge. Here, we consider a set of over 1000 time-stamped compounds, beginning with a macrocyclic natural-product lead and ending with a broad-spectrum crop anti-fungal. We demonstrate the application of the QuanSA 3D-QSAR method employing an active learning procedure that combines two types of molecular selection. The first identifies compounds predicted to be most active of those most likely to be well-covered by the model. The second identifies compounds predicted to be most informative based on exhibiting low predicted activity but showing high 3D similarity to a highly active nearest-neighbor training molecule. Beginning with just 100 compounds, using a deterministic and automatic procedure, five rounds of 20-compound selection and model refinement identifies the binding metabolic form of florylpicoxamid. We show how iterative refinement broadens the domain of applicability of the successive models while also enhancing predictive accuracy. We also demonstrate how a simple method requiring very sparse data can be used to generate relevant ideas for synthetic candidates.

C5 methylation confers accessibility, stability and selectivity to picrotoxinin.

Minor changes to complex structures can exert major influences on synthesis strategy and functional properties. Here we explore two parallel series of picrotoxinin (PXN, 1) analogs and identify leads with selectivity between mammalian and insect ion channels. These are the first SAR studies of PXN despite its >100-year history and are made possible by advances in total synthesis. We observe a remarkable stabilizing effect of a C5 methyl, which completely blocks C15 alcoholysis via destabilization of an intermediate twist-boat conformer; suppression of this secondary hydrolysis pathway increases half-life in plasma. C5 methylation also decreases potency against vertebrate ion channels (γ-Aminobutyric acid type A (GABAA) receptors) but maintains or increases antagonism of homologous invertebrate GABA-gated chloride channels (resistance to dieldrin (RDL) receptors). Optimal 5MePXN analogs appear to change the PXN binding pose within GABAARs by disruption of a hydrogen bond network. These discoveries were made possible by the lower synthetic burden of 5MePXN (2) and were illuminated by the parallel analog series, which allowed characterization of the role of the synthetically simplifying C5 methyl in channel selectivity. These are the first SAR studies to identify changes to PXN that increase the GABAA-RDL selectivity index.

Discovery of florylpicoxamid, a mimic of a macrocyclic natural product.

Natural products have routinely been used both as sources of and inspiration for new crop protection active ingredients. The natural product UK-2A has potent anti-fungal activity but lacks key attributes for field translation. Post-fermentation conversion of UK-2A to fenpicoxamid resulted in an active ingredient with a new target site of action for cereal and banana pathogens. Here we demonstrate the creation of a synthetic variant of fenpicoxamid via identification of the structural elements of UK-2A that are needed for anti-fungal activity. Florylpicoxamid is a non-macrocyclic active ingredient bearing two fewer stereocenters than fenpicoxamid, controls a broad spectrum of fungal diseases at low use rates and has a concise, scalable route which is aligned with green chemistry principles. The development of florylpicoxamid represents the first example of using a stepwise deconstruction of a macrocyclic natural product to design a fully synthetic crop protection active ingredient.

Classification of Melanocytic Lesions in Selected and Whole-Slide Images via Convolutional Neural Networks.

Whole-slide images (WSIs) are a rich new source of biomedical imaging data. The use of automated systems to classify and segment WSIs has recently come to forefront of the pathology research community. While digital slides have obvious educational and clinical uses, their most exciting potential lies in the application of quantitative computational tools to automate search tasks, assist in classic diagnostic classification tasks, and improve prognosis and theranostics. An essential step in enabling these advancements is to apply advances in machine learning and artificial intelligence from other fields to previously inaccessible pathology datasets, thereby enabling the application of new technologies to solve persistent diagnostic challenges in pathology. Here, we applied convolutional neural networks to differentiate between two forms of melanocytic lesions (Spitz and conventional). Classification accuracy at the patch level was 99.0%-2% when applied to WSI. Importantly, when the model was trained without careful image curation by a pathologist, the training took significantly longer and had lower overall performance. These results highlight the utility of augmented human intelligence in digital pathology applications, and the critical role pathologists will play in the evolution of computational pathology algorithms.

Sugar Silanes: Versatile Reagents for Stereocontrolled Glycosylation via Intramolecular Aglycone Delivery.

A new method for the intramolecular glycosylation of alcohols is described. Utilizing carbohydrate-derived silanes, the catalytic dehydrogenative silylation of alcohols is followed by intramolecular glycosylation. Appropriate combinations of silane position and protecting groups allow highly selective access to ß-manno, α-gluco, or ß-gluco stereochemical relationships as well as 2-azido-2-deoxy-ß-gluco- and 2-deoxy-ß-glucosides. Intramolecular aglycone delivery from the C-2 or C-6 position provides 1,2-cis or 1,2-trans glycosides, respectively. Multifunctional acceptor substrates such as hydroxyketones and diols are tolerated and are glycosylated in a site-selective manner.

A streamlined strategy for aglycone assembly and glycosylation.

Multipurpose sugars: Carbohydrate-derived silane reagents are utilized as the reductant for nickel-catalyzed aldehyde-alkyne reductive coupling reactions and as the glycosyl donor for subsequent intramolecular glycosylation. The approach enables the assembly of the carbon-carbon framework and stereochemical features of an aglycone while simultaneously establishing the site of glycosylation.

Ketone hydrosilylation with sugar silanes followed by intramolecular aglycone delivery: an orthogonal glycosylation strategy.

Gettin' a little sugar-no alcohol required: A procedure for the direct glycosylation of ketones without a hydroxy intermediate enables the site-selective glycosylation of hydroxyketones at the ketone or the alcohol functionality without the use of protecting groups on the aglycone (see scheme). Site selectivity is controlled by the catalyst structure in hydrosilylation and dehydrogenative silylation reactions with sugar silanes. Bn=benzyl.