3D molecular generative framework for interaction-guided drug design.

Deep generative modeling has a strong potential to accelerate drug design. However, existing generative models often face challenges in generalization due to limited data, leading to less innovative designs with often unfavorable interactions for unseen target proteins. To address these issues, we propose an interaction-aware 3D molecular generative framework that enables interaction-guided drug design inside target binding pockets. By leveraging universal patterns of protein-ligand interactions as prior knowledge, our model can achieve high generalizability with limited experimental data. Its performance has been comprehensively assessed by analyzing generated ligands for unseen targets in terms of binding pose stability, affinity, geometric patterns, diversity, and novelty. Moreover, the effective design of potential mutant-selective inhibitors demonstrates the applicability of our approach to structure-based drug design.

DFRscore: Deep Learning-Based Scoring of Synthetic Complexity with Drug-Focused Retrosynthetic Analysis for High-Throughput Virtual Screening.

Recently emerging generative AI models enable us to produce a vast number of compounds for potential applications. While they can provide novel molecular structures, the synthetic feasibility of the generated molecules is often questioned. To address this issue, a few recent studies have attempted to use deep learning models to estimate the synthetic accessibility of many molecules rapidly. However, retrosynthetic analysis tools used to train the models rely on reaction templates automatically extracted from a large reaction database that are not domain-specific and may exhibit low chemical correctness. To overcome this limitation, we introduce DFRscore (Drug-Focused Retrosynthetic score), a deep learning-based approach for a more practical assessment of synthetic accessibility in drug discovery. The DFRscore model is trained exclusively on drug-focused reactions, providing a predicted number of minimally required synthetic steps for each compound. This approach enables practitioners to filter out compounds that do not meet their desired level of synthetic accessibility at an early stage of high-throughput virtual screening for accelerated drug discovery. The proposed strategy can be easily adapted to other domains by adjusting the synthesis planning setup of the reaction templates and starting materials.

Enhancement of the solubility of recombinant proteins by fusion with a short-disordered peptide.

In protein biotechnology, large soluble fusion partners are widely utilized for increased yield and solubility of recombinant proteins. However, the production of additional large fusion partners poses an additional burden to the host, leading to a decreased protein yield. In this study, we identified two highly disordered short peptides that were able to increase the solubility of an artificially engineered aggregation-prone protein, GFP-GFIL4, from 0.6% to 61% (D3-DP00592) and 46% (D4-DP01038) selected from DisProt database. For further confirmation, the peptides were applied to two insoluble E. coli proteins (YagA and YdiU). The peptides also enhanced solubility from 52% to 90% (YagA) and from 27% to 93% (YdiU). Their ability to solubilize recombinant proteins was comparable with strong solubilizing tags, maltose-binding protein (40 kDa) and TrxA (12 kDa), but much smaller (< 7 kDa) in size. For practical application, the two peptides were fused with a restriction enzyme, I-SceI, and they increased I-SceI solubility from 24% up to 75%. The highly disordered peptides did not affect the activity of I-SceI while I-SceI fused with MBP or TrxA displayed no restriction activity. Despite the small size, the highly disordered peptides were able to solubilize recombinant proteins as efficiently as conventional fusion tags and did not interfere with the function of recombinant proteins. Consequently, the identified two highly disordered peptides would have practical utility in protein biotechnology and industry.

Biodegradation of polystyrene by bacteria from the soil in common environments.

Polystyrene (PS), a major plastic waste, is difficult to biodegrade due to its unique chemical structure that comprises phenyl moieties attached to long linear alkanes. In this study, we investigated the biodegradation of PS by mesophilic bacterial cultures obtained from various soils in common environments. Two new strains, Pseudomonas lini JNU01 and Acinetobacter johnsonii JNU01, were specifically enriched in non-carbonaceous nutrient medium, with PS as the only source of carbon. Their growth after culturing in basal media increased more than 3-fold in the presence of PS. Fourier transform infrared spectroscopy analysis, used to confirm the formation of hydroxyl groups and potentially additional chemical bond groups, showed an increase in the amount of oxidized PS samples. Moreover, field emission scanning electron microcopy analysis confirmed PS biodegradation by biofilms of the screened microbes. Water contact angle measurement additionally offered insights into the increased hydrophilic characteristics of PS films. Bioinformatics and transcriptional analysis of A. johnsonii JNU01 revealed alkane-1-monooxygenase (AlkB) to be involved in PS biodegradation, which was confirmed by the hydroxylation of PS using recombinant AlkB. These results provide significant insights into the discovery of novel functions of Pseudomonas sp. and Acinetobacter sp., as well as their potential as PS decomposers.

Learning Dynamic Textures for Neural Rendering of Human Actors.

Synthesizing realistic videos of humans using neural networks has been a popular alternative to the conventional graphics-based rendering pipeline due to its high efficiency. Existing works typically formulate this as an image-to-image translation problem in 2D screen space, which leads to artifacts such as over-smoothing, missing body parts, and temporal instability of fine-scale detail, such as pose-dependent wrinkles in the clothing. In this article, we propose a novel human video synthesis method that approaches these limiting factors by explicitly disentangling the learning of time-coherent fine-scale details from the embedding of the human in 2D screen space. More specifically, our method relies on the combination of two convolutional neural networks (CNNs). Given the pose information, the first CNN predicts a dynamic texture map that contains time-coherent high-frequency details, and the second CNN conditions the generation of the final video on the temporally coherent output of the first CNN. We demonstrate several applications of our approach, such as human reenactment and novel view synthesis from monocular video, where we show significant improvement over the state of the art both qualitatively and quantitatively.

Bifunctional Surface Coating of LiNbO3 on High-Ni Layered Cathode Materials for Lithium-Ion Batteries.

High-Ni cathode materials with a layered structure generally suffer from structural instability induced by a highly reactive Ni component, especially at the surface. Crystalline LiNbO3, with excellent thermal stability and ionic conductivity, has the potential to considerably enhance the interfacial stability of these cathode materials. By optimizing the crystalline coating of bifunctional LiNbO3 on a high-Ni cathode material, we are able to improve cycle performance and rate capability by minimizing the direct exposure of Ni with electrolytes. Since a LiNbO3 coating layer directly affects electrochemical performance, we also focus on the correlation of LiNbO3 crystallinity with electrochemical behaviors of Li+ in the cathode materials. We show that the Li+ conducting behaviors are closely related to the crystallinity of LiNbO3. Highly crystalline LiNbO3 effectively suppresses the structural changes of the cathode materials by facilitating strain relaxation induced by repeated Li+ intercalation and deintercalation into and from the host structure. Moreover, it offers strong enhancement in mechanical and thermal stabilities at elevated temperatures above 60 °C. In this regard, this research provides a practical solution for successfully utilizing high-Ni layered cathode materials in commercial LIBs.

High-Fidelity Monocular Face Reconstruction Based on an Unsupervised Model-Based Face Autoencoder.

In this work, we propose a novel model-based deep convolutional autoencoder that addresses the highly challenging problem of reconstructing a 3D human face from a single in-the-wild color image. To this end, we combine a convolutional encoder network with an expert-designed generative model that serves as decoder. The core innovation is the differentiable parametric decoder that encapsulates image formation analytically based on a generative model. Our decoder takes as input a code vector with exactly defined semantic meaning that encodes detailed face pose, shape, expression, skin reflectance, and scene illumination. Due to this new way of combining CNN-based with model-based face reconstruction, the CNN-based encoder learns to extract semantically meaningful parameters from a single monocular input image. For the first time, a CNN encoder and an expert-designed generative model can be trained end-to-end in an unsupervised manner, which renders training on very large (unlabeled) real world datasets feasible. The obtained reconstructions compare favorably to current state-of-the-art approaches in terms of quality and richness of representation. This work is an extended version of [1] , where we additionally present a stochastic vertex sampling technique for faster training of our networks, and moreover, we propose and evaluate analysis-by-synthesis and shape-from-shading refinement approaches to achieve a high-fidelity reconstruction.

Selective TiO2 Nanolayer Coating by Polydopamine Modification for Highly Stable Ni-Rich Layered Oxides.

Ni-rich layered oxides are promising cathode materials for developing high-energy lithium-ion batteries. To overcome the major challenge of surface degradation, a TiO2 surface coating based on polydopamine (PDA) modification was investigated in this study. The PDA precoating layer had abundant OH catechol groups, which attracted Ti(OEt)4 molecules in ethanol solvent and contributed towards obtaining a uniform TiO2 nanolayer after calcination. Owing to the uniform coating of the TiO2 nanolayer, TiO2 -coated PDA-LiNi0.6 Co0.2 Mn0.2 O2 (TiO2 -PNCM) displayed an excellent electrochemical stability during cycling under high voltage (3.0-4.5 V vs. Li+ /Li), during which the cathode material undergoes a highly oxidative charge process. In addition, TiO2 -PNCM exhibited excellent cyclability at elevated temperature (60 °C) compared with the bare NCM. The surface degradation of the Ni-rich cathode material, which is accelerated under harsh cycling conditions, was effectively suppressed after the formation of an ultra-thin TiO2 coating layer.

Correction: Coaxial-nanostructured MnFe2O4 nanoparticles on polydopamine-coated MWCNT for anode materials in rechargeable batteries.

Correction for 'Coaxial-nanostructured MnFe2O4 nanoparticles on polydopamine-coated MWCNT for anode materials in rechargeable batteries' by Hyeongwoo Kim et al., Nanoscale, 2018, 10, 18949-18960.

Coaxial-nanostructured MnFe2O4 nanoparticles on polydopamine-coated MWCNT for anode materials in rechargeable batteries.

MnFe2O4@PDA-coated MWCNT coaxial nanocables are successfully designed via a simple one-pot process by utilizing the adhesion property of polydopamine (PDA) with cations in aqueous solutions and employing a modified co-precipitation synthesis at a low temperature. The incorporation of the PDA coating layer on the MWCNT leads to the well-dispersed state of the MWCNTs in the aqueous solution due to the hydrophilic functional group of the PDA coating layer. In addition, the catechol-based functional group of the PDA coating layer effectively anchors the Mn and Fe ions from the aqueous solution before the co-precipitation process, eventually resulting in the preferential and homogeneous formation of MnFe2O4 nanoparticles on the MWCNT. The final MnFe2O4@PDA-coated MWCNT electrode exhibits excellent power characteristics such as a high rate capacity of around of 367 mA h g-1 at a 5C-rate condition (= 4585 mA g-1). Cycling tests reveal that the stable performance of the MnFe2O4@PDA-coated MWCNT electrode persists even after 350 cycles.

Regioselective C-H alkenylation of imidazoles and its application to the synthesis of unsymmetrically substituted benzimidazoles.

A palladium-catalyzed C-H alkenylation of imidazoles has been developed. High C5 selectivity was achieved for C2-unsubstituted and C2-substituted imidazoles using oxygen and copper(ii) acetate, respectively, as oxidants. The obtained products were applied to benzannulation through a sequence involving transposition of N-alkyl groups to give C4-alkenyl imidazoles, alkenylation, thermal 6π-electrocyclization, and oxidation, affording unsymmetrically substituted benzimidazoles.

Partial Sum Minimization of Singular Values in Robust PCA: Algorithm and Applications.

Robust Principal Component Analysis (RPCA) via rank minimization is a powerful tool for recovering underlying low-rank structure of clean data corrupted with sparse noise/outliers. In many low-level vision problems, not only it is known that the underlying structure of clean data is low-rank, but the exact rank of clean data is also known. Yet, when applying conventional rank minimization for those problems, the objective function is formulated in a way that does not fully utilize a priori target rank information about the problems. This observation motivates us to investigate whether there is a better alternative solution when using rank minimization. In this paper, instead of minimizing the nuclear norm, we propose to minimize the partial sum of singular values, which implicitly encourages the target rank constraint. Our experimental analyses show that, when the number of samples is deficient, our approach leads to a higher success rate than conventional rank minimization, while the solutions obtained by the two approaches are almost identical when the number of samples is more than sufficient. We apply our approach to various low-level vision problems, e.g., high dynamic range imaging, motion edge detection, photometric stereo, image alignment and recovery, and show that our results outperform those obtained by the conventional nuclear norm rank minimization method.

High-quality depth map upsampling and completion for RGB-D cameras.

This paper describes an application framework to perform high-quality upsampling and completion on noisy depth maps. Our framework targets a complementary system setup, which consists of a depth camera coupled with an RGB camera. Inspired by a recent work that uses a nonlocal structure regularization, we regularize depth maps in order to maintain fine details and structures. We extend this regularization by combining the additional high-resolution RGB input when upsampling a low-resolution depth map together with a weighting scheme that favors structure details. Our technique is also able to repair large holes in a depth map with consideration of structures and discontinuities utilizing edge information from the RGB input. Quantitative and qualitative results show that our method outperforms existing approaches for depth map upsampling and completion. We describe the complete process for this system, including device calibration, scene warping for input alignment, and even how our framework can be extended for video depth-map completion with the consideration of temporal coherence.

Oxidation of polyvinyl alcohol by persulfate activated with heat, Fe2+, and zero-valent iron.

The oxidation of polyvinyl alcohol (PVA) by persulfate (S(2)O(8)(2-)) activated with heat, Fe(2+), and zero-valent iron (Fe(0)) was investigated via batch experiments. It was hypothesized that elevated temperature and the addition of Fe(2+) or Fe(0) into a persulfate-water system could enhance the oxidation of PVA by activated persulfate. Increasing the temperature from 20 to 60 degrees C or 80 degrees C accelerated the oxidation rate of PVA, which achieved complete oxidation in 30 and 10 min, respectively. At 20 degrees C, the addition of Fe(2+) or Fe(0) to the persulfate-water system significantly enhanced the oxidation of PVA. The optimal persulfate-to-Fe(2+) or Fe(0) molar ratio was found to be 1:1. Complete oxidation of PVA was obtained by Fe(0)-activated persulfate in 2h. Synergistic activation of persulfate by heat and Fe(2+) or Fe(0) was also shown to enhance the oxidation of PVA in the persulfate-water system. By using GC-MS analysis, an oxidation product of PVA was identified as vinyl acetic acid (C(4)H(6)O(2)), which is readily biodegradable. Our results suggest that the oxidative treatment of PVA by activated persulfate is a viable option for the pretreatment of PVA-laden wastewater to enhance its biodegradability.