Multi-granularity adaptive extractive document summarization with heterogeneous graph neural networks.

The crucial aspect of extractive document summarization lies in understanding the interrelations between sentences. Documents inherently comprise a multitude of sentences, and sentence-level models frequently fail to consider the relationships between distantly-placed sentences, resulting in the omission of significant information in the summary. Moreover, information within documents tends to be distributed sparsely, challenging the efficacy of sentence-level models. In the realm of heterogeneous graph neural networks, it has been observed that semantic nodes with varying levels of granularity encapsulate distinct semantic connections. Initially, the incorporation of edge features into the computation of dynamic graph attention networks is performed to account for node relationships. Subsequently, given the multiplicity of topics in a document or a set of documents, a topic model is employed to extract topic-specific features and the probability distribution linking these topics with sentence nodes. Last but not least, the model defines nodes with different levels of granularity-ranging from documents and topics to sentences-and these various nodes necessitate different propagation widths and depths for capturing intricate relationships in the information being disseminated. Adaptive measures are taken to learn the importance and correlation between nodes of different granularities in terms of both width and depth. Experimental evidence from two benchmark datasets highlights the superior performance of the proposed model, as assessed by ROUGE metrics, in comparison to existing approaches, even in the absence of pre-trained language models. Additionally, an ablation study confirms the positive impact of each individual module on the model's ROUGE scores.

Identifying the roles of imine and alkyne linkages in determining the photocatalytic hydrogen evolution over thiadiazole-based covalent organic frameworks.

Covalent organic frameworks (COFs) have emerged as an extremely promising material for photocatalytic water splitting for hydrogen production. However, their photocatalytic performance is seriously affected by the properties of their donors, acceptors and linkages. So far, few studies have been reported on the key roles of the linkages of a specific COF in improving its photocatalytic hydrogen production performance. Herein, this study designed and synthesized two thiadiazole-based COFs linked by imine and alkyne bonds. The results show that the photocatalytic hydrogen production performance of imine-linked COFs (TeTz-COF1) is 19.6 times higher than that of alkyne-linked COFs (TeTz-COF2). Impressively, TeTz-COF1 achieves an apparent quantum efficiency of 3.5% at 475 nm due to the presence of imine bonds. The experimental results confirm that TeTz-COF1 with imine linkages shows higher photocurrent density, lower photocurrent resistance, and longer fluorescence lifetime than TeTz-COF2 with alkyne linkages. Meanwhile, the well-defined density functional theory (DFT) calculations further suggest that both the imine bond and the acetylene bond belong to the HOMO orbitals. Particularly, the imine bonds endow TeTz-COF1 with more delocalized orbital occupation and smaller work functions, thus leading to its lower excited state energy, stronger carrier separation ability and faster electron migration capability. Both theoretical analysis and the experimental results prove that the presence of imine bonds in TeTz-COF1 can enable the efficient separation and fast transport of photogenerated carriers and high reducing ability of photogenerated electrons. This work may provide important guiding significance for the development of new COFs in the direction of photocatalytic water splitting for hydrogen production.

Auxin regulation and MdPIN expression during adventitious root initiation in apple cuttings.

Adventitious root (AR) formation is critical for the efficient propagation of elite horticultural and forestry crops. Despite decades of research, the cellular processes and molecular mechanisms underlying AR induction in woody plants remain obscure. We examined the details of AR formation in apple (Malus domestica) M.9 rootstock, the most widely used dwarf rootstock for intensive production, and investigated the role of polar auxin transport in postembryonic organogenesis. AR formation begins with a series of founder cell divisions and elongation of the interfascicular cambium adjacent to vascular tissues. This process is associated with a relatively high indole acetic acid (IAA) content and hydrolysis of starch grains. Exogenous auxin treatment promoted this cell division, as well as the proliferation and reorganization of the endoplasmic reticulum and Golgi membrane. In contrast, treatment with the auxin transport inhibitor N-1-naphthylphthalamic acid (NPA) inhibited cell division in the basal region of the cuttings and resulted in abnormal cell divisions during the early stage of AR formation. In addition, PIN-FORMED (PIN) transcripts were differentially expressed throughout the whole AR development process. We also detected upregulation of MdPIN8 and MdPIN10 during induction; upregulation of MdPIN4, MdPIN5, and MdPIN8 during extension; and upregulation of all MdPINs during AR initiation. This research provides an improved understanding of the cellular and molecular underpinnings of the AR process in woody plants.

Complete genome sequence of Bacillus licheniformis BL-010.

The biodegradation of Aflatoxin B1 (AFB1) is an industry of increasing importance. Bacillus licheniformis BL-010 was isolated from the aflatoxin contaminated corn feed storage, and was shown to degrade AFB1 efficiently. Here we present the complete genome sequence of BL-010, the genome comprises 4,287,714 bp in a circular chromosome with a GC content of 46.12% and contains genes encoding AFB1 degrading enzymes. The genome sequence displayed that this strain contains genes involved in production of laccase, aromatic ring-opening dioxygenase which could detoxify AFB1. Complete genome sequence of the strain BL-010 can further provide the genomic basis for the biotechnological application of strain BL-010 as an effective way to degrade AFB1.

Genome-wide identification and expression analysis of the apple ASR gene family in response to Alternaria alternata f. sp. mali.

The ABA/water stress/ripening-induced (ASR) gene family exists universally in higher plants, and many ASR genes are up-regulated during periods of environmental stress and fruit ripening. Although a considerable amount of research has been performed investigating ASR gene response to abiotic stresses, relatively little is known about their roles in response to biotic stresses. In this report, we identified five ASR genes in apple (Malus × domestica) and explored their phylogenetic relationship, duplication events, and selective pressure. Five apple ASR genes (Md-ASR) were divided into two clades based on phylogenetic analysis. Species-specific duplication was detected in M. domestica ASR genes. Leaves of 'Golden delicious' and 'Starking' were infected with Alternaria alternata f. sp. mali, which causes apple blotch disease, and examined for the expression of the ASR genes in lesion areas during the first 72 h after inoculation. Md-ASR genes showed different expression patterns at different sampling times in 'Golden delicious' and 'Starking'. The activities of stress-related enzymes, peroxidase (POD), superoxide dismutase (SOD), catalase (CAT), phenylalanine ammonia lyase (PAL), and polyphenoloxidase (PPO), and the content of malondialdehyde (MDA) were also measured in different stages of disease development in two cultivars. The ASR gene expression patterns and theses physiological indexes for disease resistance suggested that Md-ASR genes are involved in biotic stress responses in apple.

Species-specific duplications of NBS-encoding genes in Chinese chestnut (Castanea mollissima).

The disease resistance (R) genes play an important role in protecting plants from infection by diverse pathogens in the environment. The nucleotide-binding site (NBS)-leucine-rich repeat (LRR) class of genes is one of the largest R gene families. Chinese chestnut (Castanea mollissima) is resistant to Chestnut Blight Disease, but relatively little is known about the resistance mechanism. We identified 519 NBS-encoding genes, including 374 NBS-LRR genes and 145 NBS-only genes. The majority of Ka/Ks were less than 1, suggesting the purifying selection operated during the evolutionary history of NBS-encoding genes. A minority (4/34) of Ka/Ks in non-TIR gene families were greater than 1, showing that some genes were under positive selection pressure. Furthermore, Ks peaked at a range of 0.4 to 0.5, indicating that ancient duplications arose during the evolution. The relationship between Ka/Ks and Ks indicated greater selective pressure on the newer and older genes with the critical value of Ks = 0.4-0.5. Notably, species-specific duplications were detected in NBS-encoding genes. In addition, the group of RPW8-NBS-encoding genes clustered together as an independent clade located at a relatively basal position in the phylogenetic tree. Many cis-acting elements related to plant defense responses were detected in promoters of NBS-encoding genes.

Determination of leucomalachite green, leucocrystal violet and their chromic forms using excitation-emission matrix fluorescence coupled with second-order calibration after dispersive liquid-liquid microextraction.

A novel spectrofluorimetric method has been developed for the simultaneous determination of leucomalachite green (LMG), leucocrystal violet (LCV), malachite green (MG) and crystal violet (CV) by combining the sensitivity of molecular fluorescence and the selectivity of the second-order calibration. Residues of LMG, LCV, MG and CV were simultaneously extracted from fish and shrimp muscle with acetonitrile. The non-fluorescent CV and MG were then reduced to the corresponding fluorescent LMG and LCV by reacting with sodium borohydride. After preconcentration with dispersive liquid-liquid microextraction technique, the extracts were analyzed by using excitation-emission matrix fluorescence coupled with second-order calibration methods based on parallel factor analysis (PARAFAC) and alternating trilinear decomposition (ATLD) algorithms. The limits of detection obtained were 2.21-2.65 ng g(-1) by PARAFAC and 2.30-2.86 ng g(-1) by ATLD, respectively. The developed method was successfully applied to simultaneous determination of the four analytes in grass carp and shrimp samples with recoveries of 90.53-103.03% for PARAFAC and 90.40-102.75% for ATLD. The accuracy of this novel method was also verified by high performance liquid chromatography.

A novel core-shell magnetic nano-sorbent with surface molecularly imprinted polymer coating for the selective solid phase extraction of dimetridazole.

A novel core-shell magnetic nano-sorbent with surface molecularly-imprinted polymer coating was prepared via a sol-gel process. Methyltrimethoxysilane and 3-aminopropyltriethoxysilane were used as functional monomers, tetraethyl orthosilicate as cross-linker, and Al(3+) as dopant to generate Lewis acid sites in the silica matrix for the metal coordinate interactions with the template dimetridazole (DMZ). The ratios of the monomers, the dopant, and the cross-linker, were optimised by a OA9 (3(4)) orthogonal array design. The resultant sorbent was characterised by scanning electron microscopy, transmission electron microscopy, Fourier transform infrared spectroscopy, powder X-ray diffraction, and magnetometry. The binding performances of the sorbent were evaluated by static, kinetic and selective adsorption experiments. The nano-sorbent was successfully applied to solid phase extraction followed by spectrophotometric determination of DMZ in real samples. Spiked recoveries ranged from 90.33% to 106.20% for egg, milk powder, and pig feed samples, with relative standard deviations of less than 4.54%.

Application of derivative and derivative ratio spectrophotometry to simultaneous trace determination of rhodamine B and rhodamine 6G after dispersive liquid-liquid microextraction.

Two novel methods, first derivative spectrophotometric method ((1)D) and first derivative ratio spectrophotometric method ((1)DR), have been developed for the simultaneous trace determination of rhodamine B (RhB) and rhodamine 6G (Rh6G) in food samples after dispersive liquid-liquid microextraction (DLLME). The combination of derivative spectrophotometric techniques and DLLME procedure endows the presented methods with enhanced sensitivity and selectivity. Under optimum conditions, the linear calibration curves ranged from 5 to 450 ng mL(-1), with the correlation coefficients (r) of 0.9997 for RhB and 0.9977 for Rh6G by (1)D method, and 0.9987 for RhB and 0.9958 for Rh6G by (1)DR method, respectively. The calculated limits of detection (LODs) based on the variability of the blank solutions (S/N = 3 criterion) for 11 measurements were in the range of 0.48-1.93 ng mL(-1). The recoveries ranged from 88.1% to 111.6% (with RSD less than 4.4%) and 91.5-110.5% (with RSD less than 4.7%) for (1)D and (1)DR method, respectively. The influence of interfering substances such as foreign ions and food colorants which might be present in the food samples on the signals of RhB and Rh6G was examined. The developed methods have been successfully applied to the determination of RhB and Rh6G in black tea, red wine and chilli powder samples with the characteristics of simplicity, cost-effectiveness, environmental friendliness, and could be valuable for routine analysis.