Explore the vast range of titles available.

Approximately 95% of ovarian teratomas are classified as benign, with only about 5% being malignant. Nevertheless, no research has previously explored the connection between primary tumor laterality and prognostic outcomes among malignant ovarian teratoma (MOT) patients. Our aim was to investigate the association of primary tumor laterality with prognosis in MOT. This retrospective study enrolled patients with MOT from The Affiliated Hospital of Qingdao University from January 2012 to December 2021. The primary outcome was progression-free survival (PFS) and overall survival (OS). The prognostic difference between left-sided, right-sided and bilateral groups was investigated using Kaplan-Meier analyses and Cox proportional hazards regression analyses. A total of 51 eligible patients with MOT were included with a median age of 52 years. Among the patients with MOT, the multivariate Cox regression analyses showed that right-sided (hazard ratio [HR]=0.01; 95% confidence interval [CI]=0.00-0.04; = 0.01) MOT was associated with better PFS, compared with left-sided MOT (HR = 1.00). Kaplan-Meier analyses also showed that the primary tumor laterality had a significant prognostic effect in MOT. Among patients with MOT, those with unilateral tumors, particularly right-sided ones, had a significantly better prognosis than those with bilateral tumors. Gynecologic oncologists might account for the prognostic impact of primary tumor laterality in MOT and tailor treatment and surveillance accordingly.

Background Pneumonia is a common respiratory disorder, which brings an enormous financial burden to the medical system. However, the current treatment options for pneumonia are limited because of drug resistance and side effects. Our previous study preliminarily confirmed that Yinlai Decoction (YD), a common prescription for pneumonia in clinical practice, can regulate the expression of inflammatory factors, but the mechanisms are unknown yet. Methods In our work, a method named network pharmacology was applied, which investigated the underlying mechanisms of herbs based on a variety of databases. We obtained bioactive ingredients of YD on TCMSP database and collected potential targets of these ingredients by target fishing. Then the pneumonia-related targets database was built by TTD, Drugbank, HPO, OMIM, and CTD. Based on the matching targets between YD and pneumonia, the PPI network was built by STRING to analyze the interactions among these targets and then input into Cytoscape for further topological analysis. DAVID and KEGG were utilized for GO and pathway enrichment analysis. Then rat model based on LPS stimulated pneumonia was used to verify the possible mechanism of YD in treating pneumonia. Results Sixty-eight active ingredients, 103 potential targets and 8 related pathways, which likely exert a number of effects, were identified. Three networks were constructed using Cytoscape, which were herb-component-network, YD-pneumonia target network, and herb-component-YD target-pneumonia network. YD was verified to treat LPS-induced pneumonia by regulating the inflammatory factor IL-6, which was a predicted target. Conclusion Network analysis indicated that YD could alleviate the symptoms and signs of pneumonia through regulating host immune inflammatory response, angiogenesis and vascular permeability, the barrier function of the airway epithelial cells, hormone releasing and cell growth, proliferation, and apoptosis.

Data analysis provides important insights to inform decision-making by discovering patterns and trends in datasets. With the increasing collection and availability of large datasets, data science has the potential to revolutionize decision-making in many fields. Basketball is one of the most popular sports in the world, and the application of data science in basketball has attracted a large amount of attention in recent years. Our research explores the dissimilarity and similarity of game-play characteristics through recent years within the National Basketball Association. We focused on the data analysis of basketball teams and player performance, analysis, and prediction of competition results. We used multidimensional scaling and autoencoder to examine the dissimilarity and similarity of team performance. The research will show that data analysis can enable teams to increase their understanding of past performance and factors that drove success. Teams can also use analysis to make predictions of future success and to inform decisions that lead to success.

In this work, we demonstrate for the first time, a method to synthesize phenylboronic acid-Fe304@polydopamine （Fe3O4@ PDA-PBA） magnetic microspheres via the combination of mussel-inspired polydopamine coating and click chemistry. Uniform-size and core-shell structured Fe3O4@PDA-PBA magnetic microspheres with a core diameter of -240 nm and a shell thickness of -13 nm were obtained as identified by the characterization of the morphology, structure and composition of the synthesized microspheres. We evaluated the selectivity and binding capacity of the Fe3O4@PDA-PBA magnetic microsphcres by using standard glycoproteins （ovalbumin, immunoglobulin G and catalase） and nonglycoproteins （human serum albumin, bovine hemoglobin, myoglobin, lysozyme, and ribonuclease A） as model proteins. Adsorption experiments, SDS-PAGE and mass spectrometry analysis demonstrated that the Fe3O4@PDA-PBA magnetic microspheres had much high binding capacity and selectivity for glycoproteins/glycopeptides compared to nonglycoproteins/nonglycopeptides. In addition, the practicability of the Fe3O4@PDA-PBA magnetic microspheres was further assessed by selective capture of glycoproteins from healthy hu- man serum. The good results demonstrated its potential in glycoproteome analysis.

To help the open-source community have a better understanding of Mixture-of-Experts (MoE) based large language models (LLMs), we train and release OpenMoE, a series of fully open-sourced and reproducible decoder-only MoE LLMs, ranging from 650M to 34B parameters and trained on up to over 1T tokens. Our investigation confirms that MoE-based LLMs can offer a more favorable cost-effectiveness trade-off than dense LLMs, highlighting the potential effectiveness for future LLM development. One more important contribution of this study is an in-depth analysis of the routing mechanisms within our OpenMoE models, leading to three significant findings: Context-Independent Specialization, Early Routing Learning, and Drop-towards-the-End. We discovered that routing decisions in MoE models are predominantly based on token IDs, with minimal context relevance. The token-to-expert assignments are determined early in the pre-training phase and remain largely unchanged. This imperfect routing can result in performance degradation, particularly in sequential tasks like multi-turn conversations, where tokens appearing later in a sequence are more likely to be dropped. Finally, we rethink our design based on the above-mentioned observations and analysis. To facilitate future MoE LLM development, we propose potential strategies for mitigating the issues we found and further improving off-the-shelf MoE LLM designs.

Perceiving and generating diverse modalities are crucial for AI models to effectively learn from and engage with real-world signals, necessitating reliable evaluations for their development. We identify two major issues in current evaluations: (1) inconsistent standards, shaped by different communities with varying protocols and maturity levels; and (2) significant query, grading, and generalization biases. To address these, we introduce MixEval-X, the first any-to-any, real-world benchmark designed to optimize and standardize evaluations across diverse input and output modalities. We propose multi-modal benchmark mixture and adaptation-rectification pipelines to reconstruct real-world task distributions, ensuring evaluations generalize effectively to real-world use cases. Extensive meta-evaluations show our approach effectively aligns benchmark samples with real-world task distributions. Meanwhile, MixEval-X's model rankings correlate strongly with that of crowd-sourced real-world evaluations (up to 0.98) while being much more efficient. We provide comprehensive leaderboards to rerank existing models and organizations and offer insights to enhance understanding of multi-modal evaluations and inform future research.

Hyperspectral images possess a continuous and analogous spectral nature, enabling the classification of distinctive information by analyzing the subtle variations between adjacent spectra. Meanwhile, a hyperspectral dataset includes redundant and noisy information in addition to larger dimensions, which is the primary barrier preventing its use for land cover categorization. Despite the excellent feature extraction capability exhibited by convolutional neural networks, its efficacy is restricted by the constrained receptive field and the inability to acquire long-range features due to the limited size of the convolutional kernels. We construct a dual-stream self-attention fusion network (DSSFN) that combines spectral and spatial information in order to achieve the deep mining of global information via a self-attention mechanism. In addition, dimensionality reduction is required to reduce redundant data and eliminate noisy bands, hence enhancing the performance of hyperspectral classification. A unique band selection algorithm is proposed in this study. This algorithm, which is based on a sliding window grouped normalized matching filter for nearby bands (SWGMF), can minimize the dimensionality of the data while preserving the corresponding spectral information. Comprehensive experiments are carried out on four well-known hyperspectral datasets, where the proposed DSSFN achieves higher classification results in terms of overall accuracy (OA), average accuracy (AA), and kappa than previous approaches. A variety of trials verify the superiority and huge potential of DSSFN.

Ti 4+ -immobilized hierarchically porous zirconium-organic frameworks (denoted as THZr-MOFs) was prepared for phosphopeptide enrichment. The THZr-MOFs showed high specific surface area of 185.28 m 2  g −1 , wide pore-size distribution of 3 ~ 20 nm, good chemical stability and excellent hydrophilicity. Introduction of hierarchical pores in MOFs not only facilitated the accessibility of phosphopeptides to the internal metal affinity sites and reduce their mass transfer resistance, but also increased the exposure sites of metal affinity interaction and binding energies of Zr and Ti elements. Benefited from these advantages, the THZr-MOFs showed high adsorption capacity (79.8 μg mg −1 ) towards standard phosphopeptide. A low detection limit (0.05 fmol μL −1 ) and high enrichment selectivity (β-casein/BSA with a molar ratio of 1:5000) were also obtained by MALDI-TOF MS. The THZr-MOFs were applied to analyze complex samples including nonfat milk, human serum, and HeLa cell lysate. In total, 1432 phosphopeptides derived from 762 phosphoproteins were identified from human HeLa cell lysate. Graphical abstract Schematic representation of the application of Ti 4+ -immobilized hierarchically porous zirconium-organic frameworks (denoted as THZr-MOFs) in high-efficiency and selective enrichment of low-abundance phosphopeptides from the tryptic digest of human HeLa cell lysate.

This paper investigates the overload controller for low-speed UAV maneuvering flight. Compared with high-speed and supersonic vehicles, low-speed UAVs are more sensitive to atmospheric disturbances during maneuvering and their signals are more sensitive. This paper combines UAV overload control theory and L1 adaptive theory to propose a longitudinal attitude control method based on overload control for low-speed UAVs. The method uses overload as the outer loop and pitch angle rate as the inner loop, and provides a reliable control scheme for maneuvering low-speed UAVs. Through simulation experiments, the controller proposed in this paper is compared with the conventional classical three-loop overload controller. The results show that the controller of this paper has faster angular rate response and stronger anti-interference capability under the same overload corresponding conditions.

Background This study is to evaluate and compare the improvement of upper airway morphology and hyoid bone position in children with Class II mandibular retrusion treated with Invisalign mandibular advancement (MA) and Twin-Block (TB) appliances, utilizing cone beam computed tomography (CBCT). Methods 32 children aged between 8 and 11.5 years old were included in this study, with an average age of 10.2 years old. These children were divided into two groups, MA and TB, with 16 children in each group. Changes in upper airway morphology and hyoid bone position before and after treatment were analyzed using CBCT. Results (1) Changes in upper airway before and after treatment: the oropharynx volume (Or-V), the oropharynx minimum cross-sectional area (Or-mCSA), the hypopharynx volume (Hy-V), and the hypopharynx minimum cross-sectional area (Hy-mCSA) in both the MA and TB groups increased after treatment, and the differences were statistically significant ( P  < 0.05) compared to pre-treatment status. (2) Changes in hyoid bone position before and after treatment: The distances between H point and third cervical vertebra (H-C3), H point and pogonion (H-RGN), H point and mandibular plane (H-MP), H point and Frankfort horizontal plane (H-FH), H and S point (H-S), and H point and palatal plane (H-PP) in both the MA and TB groups increased after treatment, and the differences were statistically significant ( P  < 0.05). Conclusion Both MA and TB appliances effectively improved the structural narrowness of the upper airway and reduced respiratory resistance, thus improving breath quality. However, MA showed more effectiveness in improving the narrowest part of the hypopharynx compared to TB. Both appliances also promoted anterior downward movement of the hyoid bone, which opens the upper airway of the oropharynx and hypopharynx and helps the upper airway morphology return to normal range.