Stereoisomers Are Not Machine Learning's Best Friends.

This study addresses the challenge of accurately identifying stereoisomers in cheminformatics, which originates from our objective to apply machine learning to predict the association constant between cyclodextrin and a guest. Identifying stereoisomers is indeed crucial for machine learning applications. Current tools offer various molecular descriptors, including their textual representation as Isomeric SMILES that can distinguish stereoisomers. However, such representation is text-based and does not have a fixed size, so a conversion is needed to make it usable to machine learning approaches. Word embedding techniques can be used to solve this problem. Mol2vec, a word embedding approach for molecules, offers such a conversion. Unfortunately, it cannot distinguish between stereoisomers due to its inability to capture the spatial configuration of molecular structures. This study proposes several approaches that use word embedding techniques to handle molecular discrimination using stereochemical information on molecules or considering Isomeric SMILES notation as a text in Natural Language Processing. Our aim is to generate a distinct vector for each unique molecule, correctly identifying stereoisomer information in cheminformatics. The proposed approaches are then compared to our original machine learning task: predicting the association constant between cyclodextrin and a guest molecule.

Photoelectrochemical properties of copper pyrovanadate (Cu2V2O7) thin films synthesized by pulsed laser deposition.

Polymorphic phases of copper pyrovanadate (α- and ß-Cu2V2O7) were synthesized by solid state reaction and the mechanisms governing the phase transitions have been highlighted by the ThermoGravimetric Analysis (TGA) and the Differential Scanning Calorimetry (DSC). The thermal evolution of the lattice parameters was determined by high temperature X-ray Diffraction revealing negative thermal expansion coefficients. The thermogravimetric analysis coupled with differential scanning calorimetry was also used to determine the optimal conditions to obtain a dense target in order to produce thin films by the Pulsed Laser Deposition (PLD) technique. Thin films elaborated under different oxygen pressures and temperatures exhibit a ß-Cu2V2O7 polycrystalline phase and their band gap indicates absorption in the visible range. These oxides can be used as photoanodes and their photoelectrochemical properties were studied for both bulk (α-Cu2V2O7) and thin films (ß-Cu2V2O7), as a function of the wavelength and/or intensity of the luminous flux. The best photocurrent efficiency was obtained under 450 nm illumination. Moreover, in the case of thin films, we have observed a linear evolution of the current density with the luminous flux. Finally, the photostability of thin films was measured and shows a reduction in the photocurrent of 8% after 1 h of measurement. This photocorrosion phenomenon was also highlighted by the elemental mapping performed on thin films by Scanning Electron Microscopy (SEM) coupled with Energy Dispersive X-ray Spectrometry (EDS).

Design of Potent Inhibitors Targeting the Main Protease of SARS-CoV-2 Using QSAR Modeling, Molecular Docking, and Molecular Dynamics Simulations.

Severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) infection is a serious global public health threat. The evolving strains of SARS-CoV-2 have reduced the effectiveness of vaccines. Therefore, antiviral drugs against SARS-CoV-2 are urgently needed. The main protease (Mpro) of SARS-CoV-2 is an extremely potent target due to its pivotal role in virus replication and low susceptibility to mutation. In the present study, a quantitative structure-activity relationship (QSAR) study was performed to design new molecules that might have higher inhibitory activity against SARS-CoV-2 Mpro. In this context, a set of 55 dihydrophenanthrene derivatives was used to build two 2D-QSAR models using the Monte Carlo optimization method and the Genetic Algorithm Multi-Linear Regression (GA-MLR) method. From the CORAL QSAR model outputs, the promoters responsible for the increase/decrease in inhibitory activity were extracted and interpreted. The promoters responsible for an increase in activity were added to the lead compound to design new molecules. The GA-MLR QSAR model was used to ensure the inhibitory activity of the designed molecules. For further validation, the designed molecules were subjected to molecular docking analysis and molecular dynamics simulations along with an absorption, distribution, metabolism, excretion, and toxicity (ADMET) analysis. The results of this study suggest that the newly designed molecules have the potential to be developed as effective drugs against SARS-CoV-2.

Recent Insights Into Electronic Performance, Magnetism and Exchange Splittings in the Cr-substituted CaO.

The first-principles computations of density functional theory are employed to characterize the structural properties, electronic structures, and ferromagnetism induced by Cr impurities in Ca1-xCrxO compounds at concentrations x = 0. 25, 0.5, and 0.75. The dynamic stability is performed by the phonon spectra calculations. The structural parameters are computed by using Wu-Cohen generalized gradient approximation, while the electronic and magnetic properties are determined by the accurate Tran-Blaha-modified Becke-Johnson exchange potential. The crystal field, direct and indirect exchange splittings were investigated to determine the origin and stability of ferromagnetic state configuration. The Ca1-xCrxO systems have right half-metallicities, which are verified by the spin polarization of 100% and the integer values of total magnetic moments. The Ca0.75Cr0.25O, Ca0.5Cr0.5O, and Ca0.25Cr0.75O are half-metallic ferromagnetic with flip-gaps of 1.495, 0.888, and 0.218 eV, respectively. Therefore, the Ca1-xCrxO materials are suitable candidates for possible applications of spin-injection in future semiconductors spintronics.

Structure of the water-splitting photocatalyst oxysulfide α-LaOInS2 and ab initio prediction of new polymorphs.

We unveil the structure and investigate the visible light water-splitting of the photocatalyst α-LaOInS2, the second polymorph in this composition. This remarkable oxysulfide exhibits rare mixed anion InS5O octahedra leading to both O-2p and S-3p hybridized with indium states in the vicinity of the Fermi level. Ab initio structure prediction shows the stability of such heteroleptic environments and points to other hypothetical polymorphs.

Pressure dependent half-metallic ferromagnetism in inverse Heusler alloy Fe2CoAl: a DFT+U calculations.

We report the electronic and magnetic properties along with the Curie temperature (T C) of the inverse full Heusler alloy (HA) Fe2CoAl obtained by using the first-principles computational method. Our calculations suggests that Fe2CoAl is a magnetic metal when treated within PBE-GGA under the applied compressive pressures. However, the implementation of electron-electron (U) (i.e., GGA+U) with varying compressive pressure (P) drastically changes the profile of the electronic structure. The application of GGA+U along with pressure induces ferromagnetic half-metallicity with an integer value of total magnetic moment ∼4.0 µ B per unit cell. The integer value is in accordance with the Slater-Pauling's rule. Here, we demonstrate the variation of semiconducting gap in the spin down channel. The band gap increases from 0.0 eV to 0.72 eV when increasing the pressure from 0 to 30 GPa. Beyond 30 GPa, the electronic band gap decreases, and it is completely diminished at 60 GPa, exhibiting metallic behaviour. The analysis of the computed results shows that the treatment of electron-electron interactions within GGA+U and the application of compressive pressure in Fe2CoAl enables d-d orbital hybridization giving rise to a half-metal ferromagnet. The T C calculated from mean field approximation (MFA) decreases up to 30 GPa and then increases linearly up to 60 GPa.

Testing the CP-correction procedure with different DFT methods on H-bonding complexes of κ-carrabiose with water molecules.

Interaction of water molecules with κ-carrabiose disaccharide, within three H-bonding complexes, was investigated. Particular interest was focused on the way with which the BSSE correction has to be performed. Two strategies were used, either performing BSSE correction during or after optimization. For this aim, several DFT-functionals (hybrid GGA and hybrid meta-GGA) and 6-31 + G* basis set were considered. The results demonstrated the uselessness of including of BSSE-CP correction during optimization for all complexes. From a structural point of view, a proper H-bonding description was obtained using the PBE0 functional for all complexes. The basis set effect on the BSSE using B3LYP functional was also investigated. The reliability of B3LYP/6-31 + G** and B3LYP/6-31++G** models for the complexes involving one or two water molecules was reported while the use of B3LYP/6-311 + G** or B3LYP/6-311++G** levels was shown to be more appropriate for larger complexes equivalent to that involving three water molecules. CP-corrected interaction energies were demonstrated to be closer to CBS-4 M interaction energies than the uncorrected ones. Graphical abstract Functional and basis set effects on BSSE.

Density functional conformational study of 2-O-sulfated 3,6 anhydro-α-D-galactose and of neo-κ- and ι-carrabiose molecules in gas phase and water.

We examined the conformational preferences of the 2-O-sulfated-3,6-α-D-anhydrogalactose (compound I) and two 1,3 linked disaccharides constituting-κ or ι-carrageenans using density functional and ab initio methods in gas phase and aqueous solution. Systematic modifications of two torsion angles leading to 324 and 144 starting geometries for the compound I and each disaccharide were used to generate adiabatic maps using B3LYP/6-31G(d). The lower energy conformers were then fully optimized using B3LYP, B3PW91 and MP2 with several basis sets. Overall, we discuss the impact of full relaxation on the energy and structure of the dominant conformations, present the performance comparison with previous molecular mechanics calculations if available, and determine whether our results are impacted, when polarization and diffuse functions are added to the 6-31G(d) basis set, or when the MP2 level of theory is used.

Alkaline-earth metal hydrides as novel host lattices for Eu(II) luminescence.

Luminescence of divalent europium has been investigated for the first time in metal hydrides. A complete solid-solution series was found for the pseudobinary system Eu(x)Sr(1-x)H(2) [a = 637.6(1) pm -12.1(3)x pm, b = 387.0(1)-6.5(2)x pm, c = 732.2(2)-10.1(4)x pm]. Europium-doped alkaline-earth hydrides Eu(x)M(1-x)H(2) (M = Ca, Sr, Ba) with a small europium concentration (x = 0.005) exhibit luminescence with maximum emission wavelengths of 764 nm (M = Ca), 728 nm (M = Sr), and 750 nm (M = Ba); i.e., the emission energy of divalent europium shows an extremely large red shift compared to the emission energies of fluorides or oxides. Theoretical calculations (LDA+U) confirm decreasing band gaps with increasing europium content of the solid solutions.

A property-matched water-soluble analogue of the benchmark ligand PPh3.

A series of sulfonated biphenylphosphanes were readily prepared from commercially available, inexpensive, and air-stable organic compounds. Of these, the trisulfonated trisbiphenylphosphane can be considered as a true water-soluble analogue of PPh(3) as the cone angle and basicity of both phosphanes are very close and result in a similar coordination mode on palladium and rhodium complexes. The catalytic performance of the trisulfonated trisbiphenylphosphane was evaluated in the aqueous hydroformylation of 1-decene and the Tsuji-Trost reaction.