Unveiling the micronutrient-immunity puzzle in inactivated COVID-19 vaccination: A comprehensive analysis of circulating micronutrient levels and humoral responses in healthy adults.

While micronutrients are crucial for immune function, their impact on humoral responses to inactivated COVID-19 vaccination remains unclear. We investigated the associations between seven key micronutrients and antibody responses in 44 healthy adults with two doses of an inactivated COVID-19 vaccine. Blood samples were collected pre-vaccination and 28 days post-booster. We measured circulating minerals (iron, zinc, copper, and selenium) and vitamins (A, D, and E) concentrations alongside antibody responses and assessed their associations using linear regression analyses. Our analysis revealed inverse associations between blood iron and zinc concentrations and anti-SARS-CoV-2 IgM antibody binding affinity (AUC for iron: ß = -258.21, p < 0.0001; zinc: ß = -17.25, p = 0.0004). Notably, antibody quality presented complex relationships. Blood selenium was positively associated (ß = 18.61, p = 0.0030), while copper/selenium ratio was inversely associated (ß = -1.36, p = 0.0055) with the neutralizing ability against SARS-CoV-2 virus at a 1:10 plasma dilution. There was no significant association between circulating micronutrient concentrations and anti-SARS-CoV-2 IgG binding affinity. These findings suggest that circulating iron, zinc, and selenium concentrations and copper/selenium ratio, may serve as potential biomarkers for both quantity (binding affinity) and quality (neutralization) of humoral responses after inactivated COVID-19 vaccination. Furthermore, they hint at the potential of pre-vaccination dietary interventions, such as selenium supplementation, to improve vaccine efficacy. However, larger, diverse studies are needed to validate these findings. This research advances the understanding of the impact of micronutrients on vaccine response, offering the potential for personalized vaccination strategies.

Economical synthesis of γ-cyclodextrin catalyzed by oriented cyclodextrin glycosyltransferase displayed on bacterial polyhydroxyalkanoate nanogranules.

BACKGROUND: The advantages of γ-cyclodextrin (γ-CD) include its high solubility, ability to form inclusion complexes with various poorly water-soluble molecules, and favorable toxicological profile; thus, γ-CD is an attractive functional excipient widely used in many industrial settings. Unfortunately, the high cost of γ-CD caused by the low activity and stability of γ-cyclodextrin glycosyltransferase (γ-CGTase) has hampered large-scale production and application. RESULTS: This study reports the in vivo one-step production of immobilized γ-CGTase decorated on the surface of polyhydroxyalkanoate (PHA) nanogranules by the N-terminal fusion of γ-CGTase to PHA synthase via a designed linker. The immobilized γ-CGTase-PHA nanogranules showed outstanding cyclization activity of 61.25 ± 3.94 U/mg (γ-CGTase protein) and hydrolysis activity of 36,273.99 ± 1892.49 U/mg, 44.74% and 18.83% higher than that of free γ-CGTase, respectively. The nanogranules also exhibited wider optimal pH (cyclization activity 7.0-9.0, hydrolysis activity 10.0-11.0) and temperature (55-60 °C) ranges and remarkable thermo- and pH-stability, expanding its utility to adapt to wider and more severe reaction conditions than the free enzyme. A high yield of CDs (22.73%) converted from starch and a high ratio (90.86%) of γ-CD in the catalysate were achieved at pH 9.0 and 50 °C for 10 h with 1 mmol/L K+, Ca2+, and Mg2+ added to the reaction system. Moreover, γ-CGTase-PHA beads can be used at least eight times, retaining 82.04% of its initial hydrolysis activity and 75.73% of its initial cyclization activity. CONCLUSIONS: This study provides a promising nanobiocatalyst for the cost-efficient production of γ-CD, which could greatly facilitate process control and economize the production cost.

Robust Semi-Supervised Traffic Sign Recognition via Self-Training and Weakly-Supervised Learning.

Traffic sign recognition is a classification problem that poses challenges for computer vision and machine learning algorithms. Although both computer vision and machine learning techniques have constantly been improved to solve this problem, the sudden rise in the number of unlabeled traffic signs has become even more challenging. Large data collation and labeling are tedious and expensive tasks that demand much time, expert knowledge, and fiscal resources to satisfy the hunger of deep neural networks. Aside from that, the problem of having unbalanced data also poses a greater challenge to computer vision and machine learning algorithms to achieve better performance. These problems raise the need to develop algorithms that can fully exploit a large amount of unlabeled data, use a small amount of labeled samples, and be robust to data imbalance to build an efficient and high-quality classifier. In this work, we propose a novel semi-supervised classification technique that is robust to small and unbalanced data. The framework integrates weakly-supervised learning and self-training with self-paced learning to generate attention maps to augment the training set and utilizes a novel pseudo-label generation and selection algorithm to generate and select pseudo-labeled samples. The method improves the performance by: (1) normalizing the class-wise confidence levels to prevent the model from ignoring hard-to-learn samples, thereby solving the imbalanced data problem; (2) jointly learning a model and optimizing pseudo-labels generated on unlabeled data; and (3) enlarging the training set to satisfy the hunger of deep learning models. Extensive evaluations on two public traffic sign recognition datasets demonstrate the effectiveness of the proposed technique and provide a potential solution for practical applications.

An experimental study on breast lesion detection and classification from ultrasound images using deep learning architectures.

BACKGROUND: Computer-aided diagnosis (CAD) in the medical field has received more and more attention in recent years. One important CAD application is to detect and classify breast lesions in ultrasound images. Traditionally, the process of CAD for breast lesions classification is mainly composed of two separated steps: i) locate the lesion region of interests (ROI); ii) classify the located region of interests (ROI) to see if they are benign or not. However, due to the complex structure of breast and the existence of noise in the ultrasound images, traditional handcrafted feature based methods usually can not achieve satisfactory result. METHODS: With the recent advance of deep learning, the performance of object detection and classification has been boosted to a great extent. In this paper, we aim to systematically evaluate the performance of several existing state-of-the-art object detection and classification methods for breast lesions CAD. To achieve that, we have collected a new dataset consisting of 579 benign and 464 malignant lesion cases with the corresponding ultrasound images manually annotated by experienced clinicians. We evaluate different deep learning architectures and conduct comprehensive experiments on our newly collected dataset. RESULTS: For the lesion regions detecting task, Single Shot MultiBox Detector with the input size as 300×300 (SSD300) achieves the best performance in terms of average precision rate (APR), average recall rate (ARR) and F1 score. For the classification task, DenseNet is more suitable for our problems. CONCLUSIONS: Our experiments reveal that better and more efficient detection and convolutional neural network (CNN) frameworks is one important factor for better performance of detecting and classification task of the breast lesion. Another significant factor for improving the performance of detecting and classification task, which is transfer learning from the large-scale annotated ImageNet to classify breast lesion.

Influence of design probe and sequence mismatches on the efficiency of fluorescent RPA.

Recombinase polymerase amplification (RPA) is an isothermal amplification technique. Because of its short detection cycle and high specificity, it has been applied in various fields. However, the design of probe on the efficiency of RPA is not well understood and the effect of sequence mismatches of oligonucleotides on the performance of RPA is rarely discussed. In this study, we found that different primers with the same probe have a slight effect on the efficiency of fluorescent RPA, and different probes with the same amplified region have a great influence on the efficiency of fluorescent RPA. We summarized the design rules of probes suitable for fluorescent RPA by analyzing the experimental data. The rule is that the best distance between fluorescent groups in the probe is 1-2 bases, and the G content should be reduced as far as possible. In addition, we verified this rule by designing a series of probes. Furthermore, we found the base mismatches of the probe had a significant effect on RPA, which can lead to false positives and can change the amplification efficiency. However, 1-3 mismatches covering the center of the primer sequence only affect the amplification efficiency of RPA, not its specificity. And with an increase in the number of primer mismatches, the efficiency of RPA will decrease accordingly. This study suggests that the efficiency of fluorescent RPA is closely related to the probe. We recommend that when designing a fluorescent probe, one must consider the presence of closely related non-targets and specific bases.

Development of propidium monoazide-recombinase polymerase amplification (PMA-RPA) assay for rapid detection of Streptococcus pyogenes and Streptococcus agalactiae.

Streptococcus pyogenes (Group A Streptococcus, GAS) and Streptococcus agalactiae (Group B Streptococcus, GBS) are common pathogens that threaten public health. In this study, a double recombinase polymerase (RPA) amplification assay was developed to rapidly detect these pathogens. Specificity tests revealed that the GAS and GBS strains were positive for speB and SIP genes, respectively. In clinical samples, the double assay performed similarly to the traditional biochemical method. The limits of detection were both ≤100 copies per reaction. In tests for simulant-contaminated samples, bacterial-culture media containing 103 CFU/mL original concentrations of S. pyogenes and S. agalactiae were positive in RPA assays after incubating for 4 h. Results can be obtained at 37 °C in 20 min. To determine whether propidium monoazide (PMA) can eliminate the influence of DNA extracted from dead cells, a bacterial suspension was treated with PMA before DNA extraction. Findings of RPA assay showed that DNA extracted from dead cells had no fluorescence signal. Therefore, the PMA-RPA assay is a promising technology for field tests and rapid point-of-care diagnosis.

Application of Differential Proteomic Analysis to Authenticate Ophiocordyceps sinensis.

Ophiocordyceps sinensis (Berk.) Sacc. is one of the most well-known fungi in traditional Chinese medicine and is attracting attention because of its nutritious and medicinal properties. The present study aimed to produce a proteomic map to identify common O. sinensis proteins. The caterpillar body and stroma of O. sinensis collected from five locations and four fungal specimens of similar appearance were examined by two-dimensional electrophoresis (2-DE). Five proteins were identified using MALDI-TOF--TOF/MS, and the 2-DE identification pattern was provided. OCS_04585 and ß-lactamase domain-containing protein, the two abundant and characteristic proteins, were separated and purified using liquid-phase isoelectric focusing. The products were high-quality materials that can be used for future protein-function studies and immunoassay development.

Enhancing the understanding of SARS-CoV-2 protein with structure and detection methods: An integrative review.

As the rapid and accurate screening of infectious diseases can provide meaningful information for outbreak prevention and control, as well as owing to the existing limitations of the polymerase chain reaction (PCR), it is imperative to have new and validated detection techniques for SARS-CoV-2. Therefore, the rationale for outlining the techniques used to detect SARS-CoV-2 proteins and performing a comprehensive comparison to serve as a practical benchmark for future identification of similar viral proteins is clear. This review highlights the urgent need to strengthen pandemic preparedness by emphasizing the importance of integrated measures. These include improved tools for pathogen characterization, optimized societal precautions, the establishment of early warning systems, and the deployment of highly sensitive diagnostics for effective surveillance, triage, and resource management. Additionally, with an improved understanding of the virus' protein structure, considerable advances in targeted detection, treatment, and prevention strategies are expected to greatly improve our ability to respond to future outbreaks.

Proteomic Analysis Reveals the Effects of Different Dietary Protein Levels on Growth and Development of Jersey-Yak.

Jersey-yak is a hybrid offspring of Jersey cattle and yak (Bos grunniens). Changing the feeding system of Jersey-yak can significantly improve its growth performance. In this study, tandem mass tag (TMT) proteomics technology was used to determine the differentially expressed proteins (DEPs) of the longissimus lumborum (LL) muscle of Jersey-yak fed different protein levels of diet. The results showed that compared with the traditional grazing feeding, the growth performance of Jersey-yaks was significantly improved by crude protein supplementation after grazing. A total of 3368 proteins were detected in these muscle samples, of which 3365 were quantified. A total of 434 DEPs were identified. Through analyses, it was found that some pathways related to muscle growth and development were significantly enriched, such as Rap1 signaling pathway, mTOR signaling pathway, and TGF-beta signaling pathway. A number of DEPs enriched in these pathways are related to muscle cell development, differentiation, and muscle development, including integrin subunit alpha 7 (ITGA7), myosin heavy chain 8 (MYH8), and collagen type XII alpha 1 chain (COL12A1). In conclusion, the results of this study provide insights into the proteomics of different feeding patterns of Jersey-yak, providing a stronger basis for further understanding the biological mechanism of hybrid varieties.

Polymorphisms within the PRKG1 Gene of Gannan Yaks and Their Association with Milk Quality Characteristics.

Yak milk, known as the "liquid gold", is a nutritious food with extensive consumption. Compared with cow milk, yak milk contains higher levels of nutrients such as dry matter, milk fat, and milk protein, which demonstrates great potential for exploitation and utilization. Protein kinase cGMP-dependent 1 (PRKG1) is an important functional molecule in the cGMP signaling pathway, and its significant influence on milk fatty acids has been discovered. The aim of this study is to explore the correlation between single nucleotide polymorphisms (SNPs) in the PRKG1 gene and the quality traits of Gannan yak milk in order to identify candidate molecular markers for Gannan yak breeding. In this study, genotyping was performed on 172 healthy, 4-5-year-old lactating Gannan yaks with similar body types, naturally grazed, and two to three parity. Three SNPs (g.404195C>T, g.404213C>T, and g.760138T>C) were detected in the PRKG1 gene of Gannan yaks, which were uniformly distributed in the yak population. Linkage disequilibrium analysis was conducted, revealing complete linkage disequilibrium between g.404195C>T and g.404213C>T. After conducting a correlation analysis between SNPs in the PRKG1 gene and milk quality in Gannan yaks, we found that PRKG1 SNPs significantly increased the content of casein, protein, and SNFs in yak milk. Among them, the TT homozygous genotype at the PRKG1 g.404195C>T loci exhibited higher casein and protein contents compared to the CC and CT genotypes (p < 0.05). The SNP g.760138T>C locus was associated with casein, protein, SNFs, and TS traits (p < 0.05). The CC genotype had higher casein and protein contents than the TT and TA genotypes (p < 0.05). However, there were no significant differences in milk fat, lactose, and acidity among the three genotypes (p > 0.05). In summary, PRKG1 gene polymorphism can serve as a candidate molecular marker for improving milk quality in Gannan yaks.

Association between Single Nucleotide Polymorphisms of PRKD1 and KCNQ3 Gene and Milk Quality Traits in Gannan Yak (Bos grunniens).

Protein kinase D1 (PRKD1) functions primarily in normal mammary cells, and the potassium voltage-gated channel subfamily Q member 3 (KCNQ3) gene plays an important role in controlling membrane potential and neuronal excitability, it has been found that this particular gene is linked to the percentage of milk fat in dairy cows. The purpose of this study was to investigate the relationship between nucleotide polymorphisms (SNPs) of PRKD1 and KCNQ3 genes and the milk quality of Gannan yak and to find molecular marker sites that may be used for milk quality breeding of Gannan yak. Three new SNPs were detected in the PRKD1 (g.283,619T>C, g.283,659C>A) and KCNQ3 gene (g.133,741T>C) of 172 Gannan lactating female yaks by Illumina yak cGPS 7K liquid-phase microarray technology. Milk composition was analyzed using a MilkoScanTM milk composition analyzer. We found that the mutations of these three loci significantly improved the lactose, milk fat, casein, protein, non-fat milk solid (SNF) content and acidity of Gannan yaks. The lactose content of the TC heterozygous genotype population at g.283,619T>C locus was significantly higher than that of the TT wild-type population (p < 0.05); the milk fat content of the CA heterozygous genotype population at g.283,659C>A locus was significantly higher than that of the CC wild-type and AA mutant populations (p < 0.05); the casein, protein and acidity of the CC mutant and TC heterozygous groups at the g.133,741T>C locus were significantly higher than those of the wild type (p < 0.05), and the SNF of the TC heterozygous group was significantly higher than that of the mutant group (p < 0.05). The results showed that PRKD1 and KCNQ3 genes could be used as candidate genes affecting the milk traits of Gannan yak.

Enzyme-linked Fab' fragment based competitive immunoassay for ovalbumin in hot-processed food.

A hen's egg is one of the most common causes of food allergy. The allergen quantitation in hot-processed food is always a difficult task, because the protein in these samples will be denatured, insoluble, and degraded. This article presents a competitive enzyme linked immunosorbent assay (ELISA) for the quantitation of ovalbumin in hot-processed food. Its recovery was improved nearly two times by the assay method as compared with previous sandwich ELISA. The heat and DL-Dithiothreitol treated ovalbumin was used as antigen for monoclonal antibody preparation. A smaller labeled antibody molecule, horseradish peroxidase (HRP) labeled Fab' fragment, was used to replace IgG in ELISA for improving sensitivity and analytical speed of the method. A binding site protection procedure was developed for Fab' fragment labeling with HRP, which prevented damage to the Fab' binding site. The combination and separation steps were efficiently completed in an affinity spin column. Based on the optimized ELISA condition, the IC50 was 1.2 µg/mL and the coefficients of variation were less than 10%.

Development of monoclonal antibodies and quantitative sandwich enzyme linked immunosorbent assay for the characteristic sialoglycoprotein of edible bird's nest.

The article presents a sandwich enzyme linked immunosorbent assay (ELISA) for identification of edible bird's nest. The characteristic sialoglycoproteins were found by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and purified by liquid-phase isoelectric focusing (LIEF). According to the analysis, the molecular weight was 106-128 kDa and the isoelectric point was ≤pH 3.0. Two anti-characteristic sialoglycoprotein monoclonal antibodies were produced. The monoclonal antibodies were examined by western-blot assay. One of the monoclonal antibody was used as coating and the other as the enzyme-labeled antibody after being coupled to horseradish peroxidase (HRP). Based on the optimized ELISA condition, the method was established with IC(50) of 1.5 ng/mL, and low cross-reactivity with various fake materials (<0.01%). ELISA provided a suitable means for screening of a large number of samples. The coefficients of variation were between 2.9% and 5.8%.

Semantic Consistent Embedding for Domain Adaptive Zero-Shot Learning.

Unsupervised domain adaptation has limitations when encountering label discrepancy between the source and target domains. While open-set domain adaptation approaches can address situations when the target domain has additional categories, these methods can only detect them but not further classify them. In this paper, we focus on a more challenging setting dubbed Domain Adaptive Zero-Shot Learning (DAZSL), which uses semantic embeddings of class tags as the bridge between seen and unseen classes to learn the classifier for recognizing all categories in the target domain when only the supervision of seen categories in the source domain is available. The main challenge of DAZSL is to perform knowledge transfer across categories and domain styles simultaneously. To this end, we propose a novel end-to-end learning mechanism dubbed Three-way Semantic Consistent Embedding (TSCE) to embed the source domain, target domain, and semantic space into a shared space. Specifically, TSCE learns domain-irrelevant categorical prototypes from the semantic embedding of class tags and uses them as the pivots of the shared space. The source domain features are aligned with the prototypes via their supervised information. On the other hand, the mutual information maximization mechanism is introduced to push the target domain features and prototypes towards each other. By this way, our approach can align domain differences between source and target images, as well as promote knowledge transfer towards unseen classes. Moreover, as there is no supervision in the target domain, the shared space may suffer from the catastrophic forgetting problem. Hence, we further propose a ranking-based embedding alignment mechanism to maintain the consistency between the semantic space and the shared space. Experimental results on both I2AwA and I2WebV clearly validate the effectiveness of our method. Code is available at https://github.com/tiggers23/TSCE-Domain-Adaptive-Zero-Shot-Learning.

Noisy Label Learning With Provable Consistency for a Wider Family of Losses.

Deep models have achieved state-of-the-art performance on a broad range of visual recognition tasks. Nevertheless, the generalization ability of deep models is seriously affected by noisy labels. Though deep learning packages have different losses, this is not transparent for users to choose consistent losses. This paper addresses the problem of how to use abundant loss functions designed for the traditional classification problem in the presence of label noise. We present a dynamic label learning (DLL) algorithm for noisy label learning and then prove that any surrogate loss function can be used for classification with noisy labels by using our proposed algorithm, with a consistency guarantee that the label noise does not ultimately hinder the search for the optimal classifier of the noise-free sample. In addition, we provide a depth theoretical analysis of our algorithm to verify the justifies' correctness and explain the powerful robustness. Finally, experimental results on synthetic and real datasets confirm the efficiency of our algorithm and the correctness of our justifies and show that our proposed algorithm significantly outperforms or is comparable to current state-of-the-art counterparts.

Antimicrobial agent chloroxylenol targets ß­catenin­mediated Wnt signaling and exerts anticancer activity in colorectal cancer.

Chloroxylenol is the active ingredient of the antibacterial agent Dettol. The anticancer effect and underlying mechanisms of this compound and other common antimicrobial agents have not been clearly elucidated. In the present study, the effects of chloroxylenol, benzalkonium chloride, benzethonium chloride, triclosan and triclocarban on ß­catenin­mediated Wnt signaling in colorectal cancer were evaluated using the SuperTOPFlash reporter assay. It was demonstrated that chloroxylenol, but not the other antimicrobial agents tested, inhibited the Wnt/ß­catenin signaling pathway by decreasing the nuclear translocation of ß­catenin and disrupting ß­catenin/T­cell factor 4 complex, which resulted in the downregulation of the Wnt target genes Axin2, Survivin and Leucine­rich G protein­coupled receptor­5. Chloroxylenol effectively inhibited the viability, proliferation, migration and invasion, and sphere formation, and induced apoptosis in HCT116 and SW480 cells. Notably, chloroxylenol attenuated the growth of colorectal cancer in the MC38 cell xenograft model and inhibited organoid formation by the patient­derived cells. Chloroxylenol also demonstrated inhibitory effects on the stemness of colorectal cancer cells. The results of the present study demonstrated that chloroxylenol could exert anti­tumor activities in colorectal cancer by targeting the Wnt/ß­catenin signaling pathway, which provided an insight into its therapeutic potential as an anticancer agent.

Comparative Study on Nutritional Characteristics and Volatile Flavor Substances of Yak Milk in Different Regions of Gannan.

This study aimed to investigate the nutritional properties of yak milk in various areas of Gannan. The milk composition analyzer, automatic amino acid analyzer, and flavor analyzer were used to detect the conventional nutrients, amino acids, and volatile flavor substances of 249 yak milks in Meiren grassland, Xiahe grassland, and Maqu grassland (hereinafter referred to as Meiren yak, Xiahe yak, and Maqu yak) in the Gannan area. The results showed that the fat content of Meiren yak milk was significantly higher than that of Maqu yak and Xiahe yak (p < 0.05). The protein content of Meiren yak milk was significantly higher than that of Xiahe yak (p < 0.05), but not significantly different from that of Maqu yak (p > 0.05). The casein content in the milk of Maqu yak was significantly higher than that of Meiren yak and Xiahe yak (p < 0.05). There was no significant difference in the lactose content of yak milk in the three regions (p > 0.05). The content of glutamic acid in the milk of Meiren yak, Xiahe yak, and Maqu yak was noticeably high, which was 1.03 g/100 g, 1.07 g/100 g, and 1.10 g/100 g, respectively. The total amino acid (TAA) content was 4.78 g/100 g, 4.87 g/100 g, and 5.0 g/100 g, respectively. The ratios of essential amino acids (EAA) and total amino acids (TAA) in the milk of Meiren yak, Xiahe yak, and Maqu yak were 42.26%, 41.27%, and 41.39%, respectively, and the ratios of essential amino acids (EAA) and nonessential amino acids (NEAA) were 73.19%, 70.28%, and 70.61%, respectively. In the yak milk samples collected from three different regions, a total of 34 volatile flavor compounds were detected, including 10 aldehydes, five esters, six ketones, four alcohols, two acids, and seven others. The main flavor substances qualitatively obtained from Meiren yak milk were ethyl acetate, n-valeraldehyde, acetic acid, heptanal, and n-hexanal. Xiahe yak milk mainly contains ethyl acetate, isoamyl alcohol, n-valeraldehyde, heptanal, and ethyl butyrate. Maqu yak milk mainly contains ethyl acetate, n-valeraldehyde, isoamyl alcohol, heptanal, ethyl butyrate, and n-hexanal. Principal component analysis showed that the flavor difference between Xiahe yak and Maqu yak was small, while the flavor difference between Xiahe yak, Maqu yak, and Meiren yak was large. The findings of this research can serve as a foundation for the future advancement and application of yak milk.

Multi-omics revealed the effects of dietary energy levels on the rumen microbiota and metabolites in yaks under house-feeding conditions.

Yak (Bos grunniens) is a unique large ruminant species in the Qinghai-Tibetan Plateau (QTP). Changing the energy levels of their rations can significantly improve their growth performance. Therefore, studying the effects of dietary energy levels on the rumen microflora and metabolites of yak is crucial for enhancing the development of the yak industry. Currently, there is a lack of understanding regarding the impact of feeding energy diets on rumen fermentation parameters, microbial functions, and metabolites. This study was designed to determine the appropriate energy level for feeding yak. Three test diets with metabolizable energy levels of 7.57 MJ/kg, 9.44 MJ/kg, and 11.9 MJ/kg were used and the concentration of volatile fatty acids (VFA) in rumen fluid was measured. The microbial communities, functions, and metabolites in yaks were studied by 16S rRNA sequencing, metagenome, and LC-MS non-targeted metabolomics to investigate the relationships among rumen fermentation parameters, microbial diversity, and metabolites. Ration energy levels significantly affect total VFA, acetate, propionate, butyrate, iso-valerate, valerate, and acetate/propionate (p < 0.05). At the phylum level, the dominant phyla in all three treatment groups were Bacteroidota, Firmicutes, and Actinobacteriota. At the genus level, the abundance of the unclassified_o__Bacteroidales, norank_f_Muribaculaceae, Lachnospiraceae_NK4A136_group, and Family _XIII_AD3011_group showed significant differences (p < 0.05) and were significantly correlated with differential metabolites screened for phosphatidylcholine [PC(16:0/0:0), PC(18:3/0:0)], uridine 3'-monophosphate, and adenosine monophosphate, etc. CAZymes family analysis showed that GHs and CEs differed significantly among the three groups. In addition, differential metabolites were mainly enriched in the pathways of lipid metabolism, nucleotide metabolism, and biosynthesis of other secondary metabolites, and the concentrations of differential metabolites were correlated with microbial abundance. In summary, this study analyzed the effects of ration energy levels on rumen microorganisms and metabolites of yaks and their relationships. The results provided a scientific basis for the selection of dietary energy for yaks in the house feeding period in the future.

Polymorphisms of CCSER1 Gene and Their Correlation with Milk Quality Traits in Gannan Yak (Bos grunniens).

Coiled-coil serine-rich protein 1 (CCSER 1) gene is a regulatory protein gene. This gene has been reported to be associated with various economic traits in large mammals in recent years. The aim of this study was to investigate the association between CCSER1 gene single nucleotide polymorphisms (SNPs) and Gannan yaks and to identify potential molecular marker loci for breeding milk quality in Gannan yaks. We genotyped 172 Gannan yaks using Illumina Yak cGPS 7K liquid microarrays and analyzed the correlation between the three SNPs loci of the CCSER1 gene and the milk qualities of Gannan yaks, including milk fat, protein and casein. It was found that mutations at the g.183,843A>G, g.222,717C>G and g.388,723G>T loci all affected the fat, protein, casein and lactose traits of Gannan yak milk to varying extents, and that the milk quality of individuals with mutant phenotypes was significantly improved. Among them, the milk fat content of AG heterozygous genotype population at g.183,843A>G locus was significantly higher than that of AA and GG genotype populations (p < 0.05); the casein and protein content of mutant GG and CG genotype populations at g.222,717C>G locus was significantly higher than that of wild-type CC genotype population (p < 0.05); and the g.388,723G>T locus of the casein and protein contents of the mutant TT genotype population were significantly higher (p < 0.05) than those of the wild-type GG genotype population. These results provide potential molecular marker sites for Gannan yak breeding.

A Convergence Path to Deep Learning on Noisy Labels.

In many real-world machine learning classification applications, the model performance based on deep neural networks (DNNs) oftentimes suffers from label noise. Various methods have been proposed in the literature to address this issue, primarily by focusing on designing noise-tolerant loss functions, cleaning label noise, and correcting the objective loss. However, the noise-tolerant loss functions face challenges when the noise level increases. This article aims to reveal a convergence path of a trained model in the presence of label noise, and here, the convergence path depicts the evolution of a trained model over epochs. We first propose a theorem to demonstrate that any surrogate loss function can be used to learn DNNs from noisy labels. Next, theories on the general convergence path for the deep models under label noise are presented and verified through a series of experiments. In addition, we design an algorithm based on the proposed theorems that make efficient corrections on the noisy labels and achieve strong robustness in the DNN models. We designed several experiments using benchmark datasets to assess noise tolerance and verify the theorems presented in this article. The comprehensive experimental results firmly confirm our theoretical results and also clearly validate the effectiveness of our method under various levels of label noise.