Wuhan Applied Science and Technology School

Abstract This paper establishes a unified framework for analyzing the far‐field convergence rates and structural stability of subsonic Euler flows with arbitrarily large vorticity and characteristic discontinuities in infinitely long nozzles. The main approach is an Euler–Lagrange transformation that flattens the discontinuities and reformulates the flow as a quasilinear elliptic equation in divergence form, combined with weighted local average energy estimates. This method addresses challenges arising from free boundaries of discontinuities and strong vorticity effects. It is shown that the convergence is governed by the slower of an intrinsic exponential decay and the boundary‐induced decay, without the smallness or convexity assumptions. This yields the first such results for flows with characteristic discontinuities. Structural stability is established for both smooth and discontinuous flows, which have linear dependence respect to the finite boundary perturbations, thereby removing the need for a small‐perturbation assumption. Both results extend naturally to incompressible flows.

Adolescent mental health is a major global concern. Many young people live with conditions that begin before adulthood and affect education and relationships. Bipolar disorder ranks as one of the most debilitating of these conditions, characterized by recurring episodes of both elevated and depressed moods. It is associated with early onset, self-harm, and suicide. In response, China has broadened its mental health policies and programs within educational institutions. National surveys show that affective disorders, including bipolar disorder, are present among Chinese school-attending children and adolescents. Yet, most school-based work focuses on depression, anxiety, and general psychological distress. This paper reviews global and Chinese evidence on bipolar disorder in youth, with attention to learning motivation, attention, and academic functioning. It examines how Chinese school mental health education addresses serious mood disorders and identifies gaps in research on bipolar disorder among Chinese students, with implications for policy and practice.

Designing new molecules with ideal properties is a critical task in drug discovery. In recent years, the accumulation of available molecular datasets has facilitated the widespread application of deep generative models in drug design. Nonetheless, a significant challenge remains in developing highperformance generative models that not only need to produce chemically valid molecular structures but also optimize the chemical properties of the generated molecules. In this study, we introduce TransFVAE, a graph Transformer-based flow Variational AutoEncoder tailored for molecular graph generation. Our approach employs VAE as the encoder and integrates a lightweight flow model as the decoder. The encoder is strategically designed to expedite the training process of the decoder, while the decoder reciprocally enhances the performance of the encoder. Unlike some existing models that only account for local node connections, our model leverages Transformer architecture in the encoder, ensuring comprehensive consideration of global information. This enables each atom to holistically interact with all other atoms, thereby enhancing molecular attribute constraint optimization in molecular optimization tasks. Validation of our model is conducted through three core tasks: molecule generation and reconstruction, latent space visualization, and molecular optimization. The results affirm the state-of-the-art performance of our model, underscoring its substantial potential in facilitating the generation of drug molecules endowed with desired properties. All source datasets and codes can be downloaded from: https://github.com/Biowust/TransFVAE.

In order to improve the prediction and topic awareness of news communication events in Microblog, an automatic detection and summarization algorithm of news communication events in Microblog based on heterogeneous network is proposed. The heat perception model of news events in Microblog was constructed by using the grid detection method involving Microblog opinion leaders, and the features of news events in Microblog were classified and extracted by combining the dynamic coefficient multiple linear regression distribution method. According to the word frequency statistics, the contents of hot topics were learned, and a large number of keywords in Microblog were extracted. The keywords were clustered by heterogeneous networks, and the feature quantities of the abstract description topics of Microblog events reflected by keywords were extracted. According to the propagation characteristics of hot topics, the semantic index fuzzy classification detection is carried out, and the similarity of Microblog texts and Microblog fever are calculated. The Single-Pass algorithm can extract the required topics from the opinion leader's Microblog, and automatically detect and abstract Microblog news communication events according to the extracted topics. The simulation results show that the automatic detection and abstract extraction of news communication events in Microblog are realized by this method, and the hot topic recognition rate is high, and it can automatically learn and adjust with the update of data stream, and at the same time, it can realize real-time detection of sudden topics.

Peroxisomal fatty acyl-CoA reductase 1 (FAR1) is a rate-limiting enzyme for ether lipid (EL) synthesis. Gene mutations in FAR1 cause a rare human disease. Furthermore, altered EL homeostasis has also been associated with various prevalent human diseases. Despite their importance in human health, the exact cellular functions of FAR1 and EL are not well-understood. Here, we report the generation and initial characterization of the first Far1 knockout (KO) mouse model. Far1 KO mice were subviable and displayed growth retardation. The adult KO male mice had smaller testes and were infertile. H&E and immunofluorescent staining showed fewer germ cells in seminiferous tubules. Round spermatids were present but no elongated spermatids or spermatozoa were observed, suggesting a spermatogenesis arrest at this stage. Large multi-nucleated giant cells (MGC) were found lining the lumen of seminiferous tubules with many of them undergoing apoptosis. The immunofluorescent signal of TEX14, an essential component of intercellular bridges (ICB) between developing germ cells, was greatly reduced and mislocalized in KO testis, suggesting the disrupted ICBs as an underlying cause of MGC formation. Integrative analysis of our total testis RNA-sequencing results and published single-cell RNA-sequencing data unveiled cell type-specific molecular alterations underlying the spermatogenesis arrest. Many genes essential for late germ cell development showed dramatic downregulation, whereas genes essential for extracellular matrix dynamics and cell-cell interactions were among the most upregulated genes. Together, this work identified the cell type-specific requirement of ELs in spermatogenesis and suggested a critical role of Far1/ELs in the formation/maintenance of ICB during meiosis.

In these years, there are dramatic development in the fake news detection field. The spread of fake news influences people's daily life, reduces the value of real news, and sometimes blemishes people's images. This phenomenon has raised our attention, so we are interested in making some effort to reduce the negative impact of it. We focus on the point that whether the users will spread the news if all the news is read by the users (from the causal inference aspect). We use negative sampling to avoid the problem that there is only positive feedback in the real-world dataset. Then we compare IPS, MF, NCF and BPR to discover the best to help us to solve this question.