NAFLDkb: A Knowledge Base and Platform for Drug Development against Nonalcoholic Fatty Liver Disease.

Nonalcoholic fatty liver disease (NAFLD) is the most common chronic liver disease with a broad spectrum of histologic manifestations. The rapidly growing prevalence and the complex pathologic mechanisms of NAFLD pose great challenges for treatment development. Despite tremendous efforts devoted to drug development, there are no FDA-approved medicines yet. Here, we present NAFLDkb, a specialized knowledge base and platform for computer-aided drug design against NAFLD. With multiperspective information curated from diverse source materials and public databases, NAFLDkb presents the associations of drug-related entities as individual knowledge graphs. Practical drug discovery tools that facilitate the utilization and expansion of NAFLDkb have also been implemented in the web interface, including chemical structure search, drug-likeness screening, knowledge-based repositioning, and research article annotation. Moreover, case studies of a knowledge graph repositioning model and a generative neural network model are presented herein, where three repositioning drug candidates and 137 novel lead-like compounds were newly established as NAFLD pharmacotherapy options reusing data records and machine learning tools in NAFLDkb, suggesting its clinical reliability and great potential in identifying novel drug-disease associations of NAFLD and generating new insights to accelerate NAFLD drug development. NAFLDkb is freely accessible at https://www.biosino.org/nafldkb and will be updated periodically with the latest findings.

Outer Macular Microvascular Supply in Retinitis Pigmentosa Examined using Optical Coherence Tomography Angiography.

PURPOSE: To determine the vessel density of the superior (SCP) and deep retinal capillary plexuses (DCP) in patients with retinitis pigmentosa (RP) using optical coherence tomography angiography (OCTA). METHODS: This was a cross-sectional study. A total of 25 eyes of 25 healthy volunteers and 30 eyes of 17 patients with RP were evaluated in this study. The integrity of the ellipsoid zone in the macular fovea was evaluated as an intact or defect using a spectral-domain OCT. Commercial spectral domain coherence tomography angiography (OCTA) was used to scan the macular region of approximately 3 × 3 mm2. The vessel density in the SCP and DCP were calculated after appropriate layer segmentation and removal of projection artifacts. The central retinal thickness (CRT) was measured with automated software. The vessel densities in the SCP and DCP were compared between different groups using SPSS. RESULTS: A total of 25 eyes of 25 healthy subjects and 30 eyes of 17 patients with RP were evaluated in the study. There was no significant difference in ages between the two groups (F = 0.065 and P=0.937). There was a significant difference in SCP and DCP between the patients with RP and healthy individuals (P < 0.001 and P < 0.001). The DCP was significantly reduced in the parafovea region between the macular intact and defect groups (P < 0.05), except in the fovea and nasal regions. After linear regression, the DCP/SCP ratio in the whole, fovea, and parafovea regions was closely related to the DCP vessel density (P < 0.05), and CRT in the fovea and parafovea was not related to the whole DCP (P=0.186 and P=0.539). CONCLUSION: The vessel density decreased in patients with RP, especially in the DCP of the parafovea region. A greater loss of capillaries in the DCP was found when the macular region was involved. The DCP/SCP ratio may be an important indicator of RP.

Computational Analysis of Naturally Occurring Aristolochic Acid Analogues and Their Biological Sources.

Aristolochic acids are known for nephrotoxicity, and implicated in multiple cancer types such as hepatocellular carcinomas demonstrated by recent studies. Natural products that are analogues to aristolochic acids have been constantly isolated from organisms; a larger chemical space of these compounds and a wider coverage of biological sources should be determined in consideration of the potential hazard of aristolochic acid analogues and the wide distribution of their biological sources in the nature. Therefore, we carried out an in silico research of naturally occurring aristolochic acid analogues and their biological sources, as a supplement to existing studies. The result shows a chemical space of 238 naturally occurring aristolochic acid analogues that are present in 175 species of biological sources including 44 traditional medicines. With the computational estimation for toxicity and the implication in hazard assessment of a biological source with the presence of aristolochic acid analogues, we propose that additional awareness should be raised to the public for avoidance of toxic species, especially those that are used as herbal medicines and easily accessible.

NPBS database: a chemical data resource with relational data between natural products and biological sources.

NPBS (Natural Products & Biological Sources) database is a chemical data resource with relational data between natural products and biological sources, manually curated from literatures of natural product researches. The relational data link a specific species and all the natural products derived from it and contrarily link a specific natural product and all the biological sources. The biological sources cover diverse species of plant, bacterial, fungal and marine organisms; the natural molecules have proper chemical structure data and computable molecular properties and all the relational data have corresponding references. NPBS database provides a wider choice of biological sources and can be used for dereplication to prevent re-isolation and re-characterization of already known natural products. Database URL: http://www.organchem.csdb.cn/scdb/NPBS.

Virtual Screening for Reactive Natural Products and Their Probable Artifacts of Solvolysis and Oxidation.

Chemically unstable natural products are prone to show their reactivity in the procedures of extraction, purification, or identification and turn into contaminants as so-called "artifacts". However, identification of artifacts requires considerable investments in technical equipment, time, and human resources. For revealing these reactive natural products and their artifacts by computational approaches, we set up a virtual screening system to seek cases in a biochemical database. The screening system is based on deep learning models of predicting the two main classifications of conversion reactions from natural products to artifacts, namely solvolysis and oxidation. A set of result data was reviewed for checking validity of the screening system, and we screened out a batch of reactive natural products and their probable artifacts. This work provides some insights into the formations of natural product artifacts, and the result data may act as warnings regarding the improper handling of biological matrixes in multicomponent extraction.

Neural machine translation of chemical nomenclature between English and Chinese.

Machine translation of chemical nomenclature has considerable application prospect in chemical text data processing between languages. However, rule based machine translation tools have to face significant complication in rule sets building, especially in translation of chemical names between English and Chinese, which are the two most used languages of chemical nomenclature in the world. We applied two types of neural networks in the task of chemical nomenclature translation between English and Chinese, and made a comparison with an existing rule based machine translation tool. The result shows that deep learning based approaches have a great chance to precede rule based translation tools in machine translation of chemical nomenclature between English and Chinese.

Crucial role of the adhesion layer on the plasmonic fluorescence enhancement.

A nanoscale layer of chromium or titanium is commonly used in plasmonic nanoantennas to firmly adhere a gold film to a glass substrate, yet the influence of this layer on the antenna performance is often ignored. As a result, the need for the use of potentially better materials is not widely recognized. Using a single aperture milled in a gold film with 120 nm diameter as a nanobench for these investigations, we present the first experimental report of the strong dependence of the plasmonic enhancement of single-molecule fluorescence on the nature of the adhesion layer. By combining fluorescence correlation spectroscopy and fluorescence lifetime measurements, we show that this structure is very sensitive to the properties of the adhesion layer, and we detail the respective contributions of excitation and emission gains to the observed enhanced fluorescence. Any increase in the absorption losses due to the adhesion layer permittivity or thickness is shown to lower the gains in both excitation and emission, which we relate to a damping of the energy coupling at the nanoaperture. With this nanobench, we demonstrate the largest enhancement factor reported to date (25×) by using a TiO(2) adhesion layer. The experimental data are supported by numerical simulations and argue for a careful consideration of the adhesion layer while designing nanoantennas for high-efficiency single-molecule analysis.

Third-harmonic generation from arrays of sub-wavelength metal apertures.

We measure third-harmonic generation (THG) from arrays of sub-wavelength metal apertures in transmission using fundamental input at 800 nm. Samples with different aperture spacings, sizes, and shapes are used. Strong angular dependence of THG is observed, with maxima located at incidence angles corresponding to extraordinary optical transmission (EOT) for the fundamental. We demonstrate an anomalous scaling of TH intensity with aperture size, where at different EOT peaks, the TH may either increase or decrease with aperture size. The aperture shape is also shown to have a strong effect on TH output.