Enhancing Video-Language Representations with Structural Spatio-Temporal Alignment.

While pre-training large-scale video-language models (VLMs) has shown remarkable potential for various downstream video-language tasks, existing VLMs can still suffer from certain commonly seen limitations, e.g., coarse-grained cross-modal aligning, under-modeling of temporal dynamics, detached video-language view. In this work, we target enhancing VLMs with a fine-grained structural spatio-temporal alignment learning method (namely Finsta). First of all, we represent the input texts and videos with fine-grained scene graph (SG) structures, both of which are further unified into a holistic SG (HSG) for bridging two modalities. Then, an SG-based framework is built, where the textual SG (TSG) is encoded with a graph Transformer, while the video dynamic SG (DSG) and the HSG are modeled with a novel recurrent graph Transformer for spatial and temporal feature propagation. A spatial-temporal Gaussian differential graph Transformer is further devised to strengthen the sense of the changes in objects across spatial and temporal dimensions. Next, based on the fine-grained structural features of TSG and DSG, we perform object-centered spatial alignment and predicate-centered temporal alignment respectively, enhancing the video-language grounding in both the spatiality and temporality. We design our method as a plug&play system, which can be integrated into existing well-trained VLMs for further representation augmentation, without training from scratch or relying on SG annotations in downstream applications. On 6 representative VL modeling tasks over 12 datasets in both standard and long-form video scenarios, Finsta consistently improves the existing 13 strong-performing VLMs persistently, and refreshes the current state-of-the-art end task performance significantly in both the fine-tuning and zero-shot settings.

Colorable Light-Scattering Device Based on Polymer-Stabilized Ion-Doped Cholesteric Liquid Crystal and an Electrochromatic Layer.

Bistable polymer-stabilized cholesteric liquid crystal (LC) devices have been extensively researched due to their energy-saving benefits. Compared to devices with merely transparent and light-scattering states, LC devices with controlled light absorption or changeable color functions are unquestionably more intriguing. In this paper, a polymer-stabilized ion-doped cholesteric LC and an electrochromic layer are used to fabricate a colorable device which can show four operating states: transparent, light-scattering, colored transparent, and colored light-scattering. The working principle and fabrication strategy are explained in detail. Based on the dielectric response of LC, the electrohydrodynamic effect of ion-doped LC, and the redox reaction of electrochromic materials, the transparent or light-scattering state and the colored or colorless state of the device can be regulated by controlling the alternating frequency and the direction of the electric field. The display performance related to the monomer, chiral dopant, and electrochromic layer is investigated. The content of monomer and chiral dopant affects the polymer network and pitch of cholesteric LC, which then affects the driving voltages and contrast ratio. The thickness of the electrochromic layer has a significant impact on the transmittance of the device's coloring and fading states. The sample with excellent operating states is obtained by optimizing the material and the construction, which can be widely applied in smart windows and energy-saving display devices.

Vision-Language Navigation With Beam-Constrained Global Normalization.

Vision-language navigation (VLN) is a challenging task, which guides an agent to navigate in a realistic environment by natural language instructions. Sequence-to-sequence modeling is one of the most prospective architectures for the task, which achieves the agent navigation goal by a sequence of moving actions. The line of work has led to the state-of-the-art performance. Recently, several studies showed that the beam-search decoding during the inference can result in promising performance, as it ranks multiple candidate trajectories by scoring each trajectory as a whole. However, the trajectory-level score might be seriously biased during ranking. The score is a simple averaging of individual unit scores of the target-sequence actions, and these unit scores could be incomparable among different trajectories since they are calculated by a local discriminant classifier. To address this problem, we propose a global normalization strategy to rescale the scores at the trajectory level. Concretely, we present two global score functions to rerank all candidates in the output beam, resulting in more comparable trajectory scores. In this way, the bias problem can be greatly alleviated. We conduct experiments on the benchmark room-to-room (R2R) dataset of VLN to verify our method, and the results show that the proposed global method is effective, providing significant performance than the corresponding baselines. Our final model can achieve competitive performance on the VLN leaderboard.

Genome-wide differences in gene expression and alternative splicing in developing embryo and endosperm, and between F1 hybrids and their parental pure lines in sorghum.

KEY MESSAGE: Developing embryo and endosperm of sorghum show substantial and multifaceted differences in gene expression and alternative splicing, which are potentially relevant to heterosis. Differential regulation of gene expression and alternative splicing (AS) are major molecular mechanisms dictating plant growth and development, as well as underpinning heterosis in F1 hybrids. Here, using deep RNA-sequencing we analyzed differences in genome-wide gene expression and AS between developing embryo and endosperm, and between F1 hybrids and their pure-line parents in sorghum. We uncover dramatic differences in both gene expression and AS between embryo and endosperm with respect to gene features and functions, which are consistent with the fundamentally different biological roles of the two tissues. Accordingly, F1 hybrids showed substantial and multifaceted differences in gene expression and AS compared with their pure-line parents, again with clear tissue specificities including extents of difference, genes involved and functional enrichments. Our results provide useful transcriptome resources as well as novel insights for further elucidation of seed yield heterosis in sorghum and related crops.

Androgen plays an important role in regulating the synthesis of pheromone in the scent gland of muskrat.

The scent (musk) gland is an organ unique to muskrats and other scent-secreting animals, and the pheromones (musk) synthesized and secreted by the scent gland play a role in chemical communication among scent-secreting animals. The musk gland is synchronized with testicular developmental changes; however, little is known regarding androgen secretion from the testis and how this regulates pheromone synthesis and the secretion of scent. To investigate the effect of androgens on the synthesis of pheromones in the musk gland, we established a muskrat castration model by surgical removal of the testis, and analyzed the histomorphology, hormone concentration, gene expression, and changes in the chemical composition of the musk gland in castration and control groups by histomorphological analysis, Enzyme-Linked ImmunoSorbent Assay (ELISA), RNA sequencing (RNA-seq), and gas chromatography-mass spectrometry (GCMS). Histomorphological analysis results showed that after castration, muskrat gland cells underwent significant atrophy (P < 0.05). Hormone measurement results showed that there was a significant decrease in serum testosterone and muskrat musk testosterone (P < 0.05) after muskrat castration. Transcriptome sequencing results showed that 510 differentially expressed transcripts (DETs) were mainly enriched in fatty acid metabolism, terpenoid backbone biosynthesis, fatty acid degradation, PPAR signaling pathway, and fatty acid biosynthesis. GCMS results showed that macrocyclic ketones, steroids, fatty acids, alcohols, and esters in musk were significantly changed (P < 0.05). In conclusion, androgens were found to play an important function in the chemical communication exchange between muskrats through regulating pheromone synthesis in musk cells. This study provides a basis for understanding the mechanism of animal communication influenced by androgens.

Communication of Uncertainty about Preliminary Evidence and the Spread of Its Inferred Misinformation during the COVID-19 Pandemic-A Weibo Case Study.

The rapid spread of preliminary scientific evidence is raising concerns on its role in producing misinformation during the COVID-19 pandemic. This research investigated how the communication of uncertainty about preliminary evidence affects the spread of its inferred misinformation in a Weibo case study. In total, 3439 Weibo posts and 10,380 reposts regarding the misinformation of pets transmitting COVID-19 were analyzed. The results showed that attitude ambiguity toward the preliminary evidence and the stage when the evidence was first released with uncertainty were associated with higher numbers of likes and retweets of misinformation posts. Our study highlights the internal sources of misinformation and revisits the contextual perspective in misinformation studies.

Study of compositions of musks in different types secreted by forest musk deer (Moschus berezovskii).

Musk is a secretion of the forest musk deer (Moschus berezovskii). Normal musk is a brown solid secretion with a light fragrance. In this study, abnormal types of musk, namely, white and black musks, were discovered during the musk collection process. Researchers have long been concerned with the components of musk. Herein, GC-MS, headspace solid-phase microextraction (HS-SPME), and nonmetric multidimensional scaling (NMDS) were used to analyze the nonpolar organic components, volatile organic components, and sample similarities among different musks, respectively. Abundant steroid hormones and proteins were also found in the musk. The steroid hormone concentrations were detected using a radioimmunoassay (RIA). Proteins in the samples were hydrolyzed and the amino acids concentrations were detected. The steroid hormone and amino acid concentrations in white musk were significantly lower than in normal and black musks (p<0.05). The components were subjected to NMDS analysis to understand the differences in components among different types of musk, with the results suggesting that white musk was different from normal and black musks.

Regulatory Roles of Peroxisomal Metabolic Pathways Involved in Musk Secretion in Muskrats.

In this study, we analyzed the main components of muskrat musk by gas chromatography-mass spectrometry, the results showed that muskrat musk contained fatty acids (29.32%), esters (31.89%), cholesterol (4.38%), cyclic ketones (16.31%), alcohols (6.42%) and other compounds, among which 9-octadecenoic acid accounted for 4.89%. We also analyzed the genes of the metabolic pathway in the scent gland at the transcriptomic level during musk-secreting and non-secreting seasons by RNA-seq (RNA sequencing). We detected 21 genes in the peroxisomal metabolic pathways, including PEX14(peroxin-14) and ACOX3(acyl-CoA oxidase), which exhibited significant differential expression between the musk-secreting season and the non-secreting season (p < 0.05). The RNA-seq results for these genes were validated by reverse transcription PCR(RT-PCR) for both seasons. In addition, we examined changes in the composition of muskrat musk from the glandular cells of scent glands cultured in vitro after RNA interference-mediated silencing of 2 differentially expressed genes, ACOX3 and HSD17B4(D-bifunctional protein, DBP). The 9-Octadecenoic acid content in muskrat musk decreased significantly following the silencing of ACOX3 and HSD17B4(D-bifunctional protein, DBP). These results suggest that peroxisomal metabolic pathways play important roles in the regulation of musk secretion in scent glands in the muskrat.

A False Trail to Follow: Differential Effects of the Facial Feedback Signals From the Upper and Lower Face on the Recognition of Micro-Expressions.

Micro-expressions, as fleeting facial expressions, are very important for judging people's true emotions, thus can provide an essential behavioral clue for lie and dangerous demeanor detection. From embodied accounts of cognition, we derived a novel hypothesis that facial feedback from upper and lower facial regions has differential effects on micro-expression recognition. This hypothesis was tested and supported across three studies. Specifically, the results of Study 1 showed that people became better judges of intense micro-expressions with a duration of 450 ms when the facial feedback from upper face was enhanced via a restricting gel. Additional results of Study 2 showed that the recognition accuracy of subtle micro-expressions was significantly impaired under all duration conditions (50, 150, 333, and 450 ms) when facial feedback from lower face was enhanced. In addition, the results of Study 3 also revealed that blocking the facial feedback of lower face, significantly boosted the recognition accuracy of subtle and intense micro-expressions under all duration conditions (150 and 450 ms). Together, these results highlight the role of facial feedback in judging the subtle movements of micro-expressions.

Sex hormones play roles in determining musk composition during the early stages of musk secretion by musk deer (Moschus berezovskii).

Musk is a secreted external hormone or information compound that is stored in musk scent glands of the males of species within the family Moschidae, such as Moschus berezovskii. The secretion of musk changes periodically during the courtship and reproduction periods, with the early stage of secretion occurring from May to July, and the maturation stage occurring from August to April of the following year. In this study, we analyzed the dynamic changes in musk components from June to April of the following year. The result showed that musk morphological character, water content, total ion chromatographic pattern, and composition undergo seasonal change. Luminescence immunoassay and radioimmunoassay analyses were performed to determine corresponding fecal hormone levels. The results showed that testosterone, estrogen, and cortisol levels in feces change on a seasonal basis, and are significantly higher in June than in other months (p < 0.01). Correlation analysis showed that the contents of four examined musk components (muscone, cyclopentadecanone, cholesterol, and cholestenol) from June to August were significantly highly negatively correlated with fecal testosterone and estradiol levels (p < 0.01). In contrast, the correlation coefficients were low or not significant from August to April of the following year. These results indicate that testosterone and estradiol may play a major role in determining musk composition during the early stage of musk secretion but not during the course of musk maturation, which suggests that musk secretion may be promoted by increases in sex hormones in June.

The chromatin remodeler ZmCHB101 impacts expression of osmotic stress-responsive genes in maize.

KEY MESSAGE: The maize chromatin remodeler ZmCHB101 plays an essential role in the osmotic stress response. ZmCHB101 controls nucleosome densities around transcription start sites of essential stress-responsive genes. Drought and osmotic stresses are recurring conditions that severely constrain crop production. Evidence accumulated in the model plant Arabidopsis thaliana suggests that core components of SWI/SNF chromatin remodeling complexes play essential roles in abiotic stress responses. However, how maize SWI/SNF chromatin remodeling complexes function in osmotic and drought stress responses remains unknown. Here we show that ZmCHB101, a homolog of A. thaliana SWI3D in maize, plays essential roles in osmotic and dehydration stress responses. ZmCHB101-RNA interference (RNAi) transgenic plants displayed osmotic, salt and drought stress-sensitive phenotypes. Genome-wide RNA-sequencing analysis revealed that ZmCHB101 impacts the transcriptional expression landscape of osmotic stress-responsive genes. Intriguingly, ZmCHB101 controls nucleosome densities around transcription start sites of essential stress-responsive genes. Furthermore, we identified that ZmCHB101 associates with RNA polymerase II (RNAPII) in vivo and is a prerequisite for the proper occupancy of RNAPII on the proximal regions of transcription start sites of stress-response genes. Taken together, our findings suggest that ZmCHB101 affects gene expression by remodeling chromatin states and controls RNAPII occupancies in maize under osmotic stress.

Imprinted gene expression in maize starchy endosperm and aleurone tissues of reciprocal F1 hybrids at a defined developmental stage.

Imprinted gene expression in flowering plants predominantly occurs in the triploid endosperm of developing seed. However, endosperm is composed of distinct tissue types. For example, the maize (Zea mays) endosperm is constituted by two major tissues, starchy endosperm and aleurone. Previous studies in imprinted gene expression have generally assumed that the different tissues constituting endosperm would behavior the same, and hence have not examined them separately. Here, to examine parental-specific expression of imprinted genes in different parts of the seed, eight previously reported maize protein-coding imprinted genes were selected, and analyzed by cleaved amplified polymorphic sequence (CAPS) coupled with Sanger sequencing for transcripts from the various seed tissues collected at 18 days after pollination (DAP). The studied tissues included seed coat, embryo, starchy endosperm and aleurone, which were collected from a pair of reciprocal F1 hybrids produced by crossing inbred lines B73 and Mo17. Six of these eight analyzed imprinted genes showed the same imprinted expression pattern between the starchy endosperm and aleurone, but two showed imprinted expression only in the starchy endosperm. Comparison of the expression pattern of 20 selected imprinted genes in multiple seed tissues and vegetative tissues indicated that the majority (~ 75%) of these imprinted genes exhibited seed-specific or endosperm-specific expression. Our results also uncovered that imprinted genes have a high propensity to be alternatively spliced via intron retention in the developing embryo compared with the other tissues.

Comparative Analysis of Gut Microbiota Changes in Père David's Deer Populations in Beijing Milu Park and Shishou, Hubei Province in China.

This study used 16S rRNA high-throughput sequencing technology to examine the differences in gut microbiota between the Père David's deer populations in the Beijing and Shishou areas of China in order to understand the effects of ex situ conservation on the intestinal microflora in the Père David's deer. Results: On the phylum level, the main bacteria found in the Père David's deer populations from both areas were similar: Firmicutes and Bacteroidetes. However, the relative abundances of the two groups were significantly different. Alpha diversity results indicated that there was a difference in the evenness of the microflora between the two groups, and the beta diversity results further indicated that there was a significant difference in the microflora structure between the two groups. Conclusions: During the ex situ conservation process of the Père David's deer, their food sources may change, resulting in differences in the gut microbiota. The intestinal microflora in the Père David's deer from the same area are clustered. Therefore, the impact of changes in food on the gut microbiota of the Père David's deer should be taken into consideration during ex situ conservation.

Polysaccharides Extracted from Rhizoma Pleionis Have Antitumor Properties In Vitro and in an H22 Mouse Hepatoma Ascites Model In Vivo.

Malignant ascites is a highly severe and intractable complication of advanced or recurrent malignant tumors that is often immunotherapy-resistant. Rhizoma Pleionis is widely used in traditional medicine as an antimicrobial and anticancer agent, but its effectiveness in treating malignant ascites is unclear. In the current study, we investigated the effect of polysaccharides isolated from Rhizoma Pleionis (PRP) on murine hepatocarcinoma H22 cells in an ascites model. We have found that the main components of PRP, that presented a relative molecular weight of 383.57 kDa, were mannose and glucose. We also found that PRP reduced the occurrence of abdominal ascites and increased survival in our mouse model. An immune response in the ascites tumor model was observed by performing a lymphocytes proliferation experiment and an E-rosette test. The ratios of CD8+ cytotoxic T cells and NK cells in the spleen were examined by flow cytometry, and the mRNA expression of Foxp3+in CD4⁺CD25⁺ (T regulatory Tregs) was measured by RT-PCR (reverse transcription-polymerase chain reaction). The levels of the cytokines TNF-α (tumor necrosis factor), VEGF (vascular endothelial growth factor), IL-2 (interleukin), and IFN-γ (interferon) in the serum and ascites supernatants were measured by ELISA. The expression of Foxp3 and Stat3 in peritoneal cells in the mouse model was measured by immunocytochemistry. The results indicated that PRP increased H22 tumor cell apoptosis in vivo by activating and enhancing the immune response. Furthermore, the effects of PRP on the proliferation of H22 cells were assessed by the CCK8 assay, Hoechest 33258, and TUNEL staining in vitro. We found that PRP suppressed the proliferation of H22 tumor cells but had no effect on BRL (Big rat liver) -3A rat hepatoma normal cells in vitro. Next, we investigated the underlying immunological mechanism by which PRP inhibits malignant ascites. PRP induced tumor cell apoptosis by inhibiting the Jak1⁻Stat3 pathway and by activating Caspase-3 and Caspase-8 to increase the Bax/Bcl-2 ratio. Collectively, our results indicate that PRP exhibits significant antitumor properties in H22 cells in vivo and in vitro, indicating that PRP may be used as a new therapeutic drug for cancer treatment.

Comparison Between the Fecal Bacterial Microbiota of Healthy and Diarrheic Captive Musk Deer.

Diarrhea constitutes one of the most common diseases affecting the survival of captive musk deer and is usually caused by an imbalance in intestinal microbiota. Currently, research regarding the structure and function of intestinal microbiota in diarrheic musk deer is lacking. Therefore, in the present study, high-throughput 16S-rRNA gene sequencing was used to analyze the intestinal microbiota in feces of healthy captive musk deer (HMD) (n = 8) and musk deer with mild (MMD) (n = 8), and severe (n = 5) (SMD) diarrhea to compare the difference in intestinal microbiota of musk deer under various physiological conditions. The results showed that the diversity of HMD fecal microbiota was significantly higher than that of the two diarrhea samples. ß Diversity results indicated that there were extremely significant differences in bacterial communities between the HMD sample and the MMD and SMD samples. However, no significant difference was found between the two diarrhea samples. LefSe analysis showed that the degree of intestinal physiological dysfunction in musk deer was correlated with the types of major pathogens. The main pathogen in the MMD group is Escherichia-Shigella, whereas Fusobacterium is the main pathogen in the SMD group. PICRUSt functional profile prediction indicated that the intestinal microbiota disorder could also lead to changes in the abundance of genes in metabolic pathways of the immune system. Altogether, this study provides a theoretical basis for the exploration of treatments for diarrhea in captive musk deer, which is of considerable significance to the implementation of the musk deer release into the wild program.

Microbiota Changes in the Musk Gland of Male Forest Musk Deer During Musk Maturation.

The musk gland in an adult male forest musk deer is an organ that synthesizes, stores, and secretes musk, a cream-colored liquid upon initial secretion that gradually transforms into a blackish-brown solid substance upon full maturation. In this study, four healthy adult male forest musk deer were selected and a total of 12 musk samples were collected for analysis. The samples were in three different states depending on the different seasonal collection dates, which were in June, August, and October. High-throughput 16S-rRNA gene sequencing technology was used to detect microbiota changes in the gland. The results indicate that microbial richness gradually declined during the musk maturation process. The microbiota composition between the initial liquid and final solid musk samples was varied significantly (P < 0.05). The dominant bacterial phyla were similar at all three stages included Firmicutes, Proteobacteria, Actinobacteria, and Bacteroidetes. However, the abundances were differences in terms of the dominant bacterial genera. PICRUSt analysis showed the highest represented category was "Amino acid transport and metabolism" (24.8%), followed by "Transcription" (22.04%), and "Carbohydrate transport and metabolism" (20.74%). Our findings indicate that the microbiota in the musk gland plays an important role in the maturation process of musk.

Comparative Analysis of the Gut Microbiota Composition between Captive and Wild Forest Musk Deer.

The large and complex gut microbiota in animals has profound effects on feed utilization and metabolism. Currently, gastrointestinal diseases due to dysregulated gut microbiota are considered important factors that limit growth of the captive forest musk deer population. Compared with captive forest musk deer, wild forest musk deer have a wider feeding range with no dietary limitations, and their gut microbiota are in a relatively natural state. However, no reports have compared the gut microbiota between wild and captive forest musk deer. To gain insight into the composition of gut microbiota in forest musk deer under different food-source conditions, we employed high-throughput 16S rRNA sequencing technology to investigate differences in the gut microbiota occurring between captive and wild forest musk deer. Both captive and wild forest musk deer showed similar microbiota at the phylum level, which consisted mainly of Firmicutes and Bacteroidetes, although significant differences were found in their relative abundances between both groups. α-Diversity results showed that no significant differences occurred in the microbiota between both groups, while ß-diversity results showed that significant differences did occur in their microbiota compositions. In summary, our results provide important information for improving feed preparation for captive forest musk deer and implementing projects where captive forest musk deer are released into the wild.

Transcriptome analysis of muskrat scented glands degeneration mechanism.

The scented gland, a musk-secreting organ of male muskrats, shows clear seasonal changes. When entering the secreting season in March, scented glands gradually increase in size and active secretion starts. In September, scented glands become gradually smaller and secretion decreases. By November, scented glands are gradually replaced by adipose tissue. In this study, six healthy adult male muskrats were analysed: three from the secreting season (March) and three from the non-secreting season (November). Using RNA-Seq analysis, gene expression profiles of scented glands from both seasons were determined. Using the Kyoto Encyclopedia of Genes and Genomes (KEGG) database, we found that genes involved in calcium and TGF-beta signalling pathways were significantly more expressed in the non-secreting than in the secreting season. These changes in gene expression correlated with alterations in scented gland size. Both calcium and TGF-beta signalling pathways are important regulators of cell apoptosis, which may thus be involved in muskrat scented gland degeneration.

A neural joint model for entity and relation extraction from biomedical text.

BACKGROUND: Extracting biomedical entities and their relations from text has important applications on biomedical research. Previous work primarily utilized feature-based pipeline models to process this task. Many efforts need to be made on feature engineering when feature-based models are employed. Moreover, pipeline models may suffer error propagation and are not able to utilize the interactions between subtasks. Therefore, we propose a neural joint model to extract biomedical entities as well as their relations simultaneously, and it can alleviate the problems above. RESULTS: Our model was evaluated on two tasks, i.e., the task of extracting adverse drug events between drug and disease entities, and the task of extracting resident relations between bacteria and location entities. Compared with the state-of-the-art systems in these tasks, our model improved the F1 scores of the first task by 5.1% in entity recognition and 8.0% in relation extraction, and that of the second task by 9.2% in relation extraction. CONCLUSIONS: The proposed model achieves competitive performances with less work on feature engineering. We demonstrate that the model based on neural networks is effective for biomedical entity and relation extraction. In addition, parameter sharing is an alternative method for neural models to jointly process this task. Our work can facilitate the research on biomedical text mining.

Comparison of amino acid profiles and metabolic gene expression in muskrat scented glands in secretion and non-secretion season.

The scented gland is an organ responsible for producing musk in muskrats. During musk secretion season, the metabolism of glandular cells increases in the scented glands and a large amount of musk is synthesised. In this study, we collected scented gland arterial blood from six healthy adult male muskrats during non-secretion season (November). We also obtained scented gland arterial blood, venous blood, and musk from six healthy adult males during secretion season (March). Qualitative and quantitative analyses of free amino acids in blood and musk were performed with an automated amino acid analyzer. Additionally, we employed RNA sequencing technology to study the expression patterns of amino acid metabolic pathways in scented glands. Amino acid profile analysis indicates that scented glands can concentrate amino acids during secretion season, and transcriptome analysis suggests that some amino acid metabolism-related genes undergo significant seasonal changes. In summary, scented gland amino acid metabolism displays seasonal differences. Elevated amino acid metabolic activity during secretion season sustains the glands' secretory function.