Predicting retrosynthetic pathways using transformer-based models and a hyper-graph exploration strategy

We present an extension of our Molecular Transformer model combined with a hyper-graph exploration strategy for automatic retrosynthesis route planning without human intervention. The single-step retrosynthetic model sets a new state of the art for predicting reactants as well as reagents, solvents and catalysts for each retrosynthetic step. We introduce four metrics (coverage, class diversity, round-trip accuracy and Jensen–Shannon divergence) to evaluate the single-step retrosynthetic models, using the forward prediction and a reaction classification model always based on the transformer architecture. The hypergraph is constructed on the fly, and the nodes are filtered and further expanded based on a Bayesian-like probability. We critically assessed the end-to-end framework with several retrosynthesis examples from literature and academic exams. Overall, the frameworks have an excellent performance with few weaknesses related to the training data. The use of the introduced metrics opens up the possibility to optimize entire retrosynthetic frameworks by focusing on the performance of the single-step model only.

We present an extension of our Molecular Transformer model combined with a hyper-graph exploration strategy for automatic retrosynthesis route planning without human intervention.     

A Direct NMR Method To Measure Self-Diffusion Coefficients in Liquids by Monitoring Diffusing Molecules through One-Dimensional Imaging

Diffusion processes can be followed directly by recording one-dimensional images of a selected slice at variable intervals after selective inversion of the magnetization. The resulting diffusion coefficients of H₂O and DMSO are consistent with earlier studies at different temperatures, obtained by monitoring the attenuation of NMR signals as a function of the gradient amplitude in gradient echo sequences.     

Characterization of polyoxometalate/polymer photo-composites: A toolbox for the photodegradation of organic pollutants

Different inorganic/organic photocomposites based on polyoxometalate (POM) nanoparticles have been developed for photocatalytic applications. Currently, polyoxometalate nanoparticles have been successfully in-situ embedded into an acrylate polymer network by photopolymerization upon mild visible light irradiation at 405 nm. The proposed POM/polymer photocomposites have been characterized using complementary techniques for a better understanding of their photocatalytic activity. Interestingly, the obtained photocomposites exhibit high rigidity, excellent thermal stability, a non-negligible porosity and new functionalities such as light reactivity and redox properties. Moreover, developed composites showed efficient catalytic activity for the color removal of aqueous solutions of erythrosine and rose Bengal under Light Emitting Diodes LED@375 nm irradiation reaching 80 and 90% as a final color removal, respectively.     

On the Effect of the Various Assumptions and Approximations used in Molecular Simulations on the Properties of Bio-Molecular Systems: Overview and Perspective on Issues

Computer simulations of molecular systems enable structure-energy-function relationships of molecular processes to be described at the sub-atomic, atomic, supra-atomic or supra-molecular level and plays an increasingly important role in chemistry, biology and physics. To interpret the results of such simulations appropriately, the degree of uncertainty and potential errors affecting the calculated properties must be considered. Uncertainty and errors arise from (1) assumptions underlying the molecular model, force field and simulation algorithms, (2) approximations implicit in the interatomic interaction function (force field), or when integrating the equations of motion, (3) the chosen values of the parameters that determine the accuracy of the approximations used, and (4) the nature of the system and the property of interest. In this overview, advantages and shortcomings of assumptions and approximations commonly used when simulating bio-molecular systems are considered. What the developers of bio-molecular force fields and simulation software can do to facilitate and broaden research involving bio-molecular simulations is also discussed.     

Rains of solitons in a figure-of-eight passively mode-locked fiber laser

We report experimental observation of rains of solitons in figure-of-eight fiber laser passively mode-locked through nonlinear optical loop mirror. Soliton pulses are created from an extended noisy background and drift until they reach a condensed phase comprising several tens of aggregated solitons. The observation of this dynamics tends to strengthen the idea of the universality of the collective behavior of solitons.     

Application of SSA thermal fractionation and X‐ray diffraction to elucidate comonomer inclusion or exclusion from the crystalline phases in poly(butylene succinate‐ran‐butylene azelate) random copolymers

Poly(butylene succinate‐ran‐butylene azelate) random copolyesters were thermally fractionated by successive self‐nucleation and annealing (SSA). The samples before and after SSA were analyzed by differential scanning calorimetry (DSC) and X‐ray diffraction (WAXS and SAXS). WAXS results indicate that a small degree of comonomer inclusion is present in the crystalline phases that are formed in the copolymers depending on composition: a PBS‐like unit cell or/and a PBAz‐like unit cell, thus confirming the isodimorphic behavior of the samples. SSA on the other hand demonstrated that the degree of comonomer exclusion during crystallization is far larger than comonomer inclusion, as judged by the increase in fractionation degree with compositions leading to the pseudo‐eutectic point. Furthermore, WAXS, SAXS, and SSA results show that the isodimorphic behavior is not highly dependent on kinetic factors, as the degree of comonomer inclusion or exclusion in the samples was not significantly altered by SSA thermal fractionation, a thermal treatment that promotes annealing and molecular segregation of defects to the amorphous regions of the material. © 2016 Wiley Periodicals, Inc. J. Polym. Sci., Part B: Polym. Phys. 2016 , 54, 2346–2358     

Crystal structure of an antiviral ankyrin targeting the HIV-1 capsid and molecular modeling of the ankyrin-capsid complex

Ankyrins are cellular repeat proteins, which can be genetically modified to randomize amino-acid residues located at defined positions in each repeat unit, and thus create a potential binding surface adaptable to macromolecular ligands. From a phage-display library of artificial ankyrins, we have isolated Ank^GAG1D4, a trimodular ankyrin which binds to the HIV-1 capsid protein N-terminal domain (NTD^CA) and has an antiviral effect at the late steps of the virus life cycle. In this study, the determinants of the Ank^GAG1D4-NTD^CA interaction were analyzed using peptide scanning in competition ELISA, capsid mutagenesis, ankyrin crystallography and molecular modeling. We determined the Ank^GAG1D4 structure at 2.2 Å resolution, and used the crystal structure in molecular docking with a homology model of HIV-1 capsid. Our results indicated that NTD^CA alpha-helices H1 and H7 could mediate the formation of the capsid-Ank^GAG1D4 binary complex, but the interaction involving H7 was predicted to be more stable than with H1. Arginine-18 (R18) in H1, and R132 and R143 in H7 were found to be the key players of the Ank^GAG1D4-NTD^CA interaction. This was confirmed by R-to-A mutagenesis of NTD^CA, and by sequence analysis of trimodular ankyrins negative for capsid binding. In Ank^GAG1D4, major interactors common to H1 and H7 were found to be S45, Y56, R89, K122 and K123. Collectively, our ankyrin-capsid binding analysis implied a significant degree of flexibility within the NTD^CA domain of the HIV-1 capsid protein, and provided some clues for the design of new antivirals targeting the capsid protein and viral assembly.     

Strong,Thermally Superinsulating Biopolymer–Silica Aerogel Hybrids by Cogelation of Silicic Acid with Pectin

Silica aerogels are excellent thermal insulators, but their brittle nature has prevented widespread application. To overcome these mechanical limitations, silica–biopolymer hybrids are a promising alternative. A one‐pot process to monolithic, superinsulating pectin–silica hybrid aerogels is presented. Their structural and physical properties can be tuned by adjusting the gelation pH and pectin concentration. Hybrid aerogels made at pH 1.5 exhibit minimal dust release and vastly improved mechanical properties while remaining excellent thermal insulators. The change in the mechanical properties is directly linked to the observed “neck‐free” nanoscale network structure with thicker struts. Such a design is superior to “neck‐limited”, classical inorganic aerogels. This new class of materials opens up new perspectives for novel silica–biopolymer nanocomposite aerogels.     

Automated solvent system screening for the preparative countercurrent chromatography of pharmaceutical discovery compounds

A fully automated countercurrent chromatography system has been constructed to rapidly screen the commonly used heptane/ethyl acetate/methanol/water solvent system series and translate the results to preparative scale separations. The system utilizes “on‐demand” preparation of the heptane/ethyl acetate/methanol/water solvent system upper and lower phases. Elution‐extrusion countercurrent chromatography was combined with non‐dynamic equilibrium injection reducing the screening time for each heptane/ethyl acetate/methanol/water system to 17 min. The result enabled solvent system development to be reduced to under 2 h. The countercurrent chromatography system was interfaced with a mass spectrometer to allow selective detection of target components in crude medicinal chemistry reaction mixtures. Mass‐directed preparative countercurrent chromatography purification was demonstrated for the first time using a synthetic tetrazole epoxide derived from a routine medicinal chemistry support workflow.