Multiscale and Auto-Tuned Semi-Supervised Deep Subspace Clustering and Its Application in Brain Tumor Clustering

In this paper,we introduce a novel Multi-scale and Auto-tuned Semi-supervised Deep Subspace Clustering(MAS-DSC)algorithm,aimed at addressing the challenges of deep subspace clustering in high-dimensional real-world data,particularly in the field of medical imaging.Traditional deep subspace clustering algorithms,which are mostly unsupervised,are limited in their ability to effectively utilize the inherent prior knowledge in medical images.Our MAS-DSC algorithm incorporates a semi-supervised learning framework that uses a small amount of labeled data to guide the clustering process,thereby enhancing the discriminative power of the feature representations.Additionally,the multi-scale feature extraction mechanism is designed to adapt to the complexity of medical imaging data,resulting in more accurate clustering performance.To address the difficulty of hyperparameter selection in deep subspace clustering,this paper employs a Bayesian optimization algorithm for adaptive tuning of hyperparameters related to subspace clustering,prior knowledge constraints,and model loss weights.Extensive experiments on standard clustering datasets,including ORL,Coil20,and Coil100,validate the effectiveness of the MAS-DSC algorithm.The results show that with its multi-scale network structure and Bayesian hyperparameter optimization,MAS-DSC achieves excellent clustering results on these datasets.Furthermore,tests on a brain tumor dataset demonstrate the robustness of the algorithm and its ability to leverage prior knowledge for efficient feature extraction and enhanced clustering performance within a semi-supervised learning framework.

Large-scale loss-of-function perturbations reveal a comprehensive epigenetic regulatory network in breast cancer

Objective:Epigenetic abnormalities have a critical role in breast cancer by regulating gene expression;however,the intricate interrelationships and key roles of approximately 400 epigenetic regulators in breast cancer remain elusive.It is important to decipher the comprehensive epigenetic regulatory network in breast cancer cells to identify master epigenetic regulators and potential therapeutic targets.Methods:We employed high-throughput sequencing-based high-throughput screening(HTS^(2))to effectively detect changes in the expression of 2,986 genes following the knockdown of 400 epigenetic regulators.Then,bioinformatics analysis tools were used for the resulting gene expression signatures to investigate the epigenetic regulations in breast cancer.Results:Utilizing these gene expression signatures,we classified the epigenetic regulators into five distinct clusters,each characterized by specific functions.We discovered functional similarities between BAZ2B and SETMAR,as well as CLOCK and CBX3.Moreover,we observed that CLOCK functions in a manner opposite to that of HDAC8 in downstream gene regulation.Notably,we constructed an epigenetic regulatory network based on the gene expression signatures,which revealed 8 distinct modules and identified 10 master epigenetic regulators in breast cancer.Conclusions:Our work deciphered the extensive regulation among hundreds of epigenetic regulators.The identification of 10 master epigenetic regulators offers promising therapeutic targets for breast cancer treatment.

An Improved Soft Subspace Clustering Algorithm for Brain MR Image Segmentation

In recent years,the soft subspace clustering algorithm has shown good results for high-dimensional data,which can assign different weights to each cluster class and use weights to measure the contribution of each dimension in various features.The enhanced soft subspace clustering algorithm combines interclass separation and intraclass tightness information,which has strong results for image segmentation,but the clustering algorithm is vulnerable to noisy data and dependence on the initialized clustering center.However,the clustering algorithmis susceptible to the influence of noisydata and reliance on initializedclustering centers andfalls into a local optimum;the clustering effect is poor for brain MR images with unclear boundaries and noise effects.To address these problems,a soft subspace clustering algorithm for brain MR images based on genetic algorithm optimization is proposed,which combines the generalized noise technique,relaxes the equational weight constraint in the objective function as the boundary constraint,and uses a genetic algorithm as a method to optimize the initialized clustering center.The genetic algorithm finds the best clustering center and reduces the algorithm’s dependence on the initial clustering center.The experiment verifies the robustness of the algorithm,as well as the noise immunity in various ways and shows good results on the common dataset and the brain MR images provided by the Changshu First People’s Hospital with specific high accuracy for clinical medicine.

The immunoreactivity of Cra a 4 decreased after Maillard reaction with xylose and glucose

Cra a 4 is a heat stable allergen in oyster,the effect of Maillard reaction on the immunoreactivity and structure of Cra a 4 was investigated.Firstly,Cra a 4 was cloned,expressed and identified.The purified Cra a 4 was incubated with xylose at 100℃,pH 8.5 for 30 min or with glucose for 60 min.After Maillard reaction,the IgE-binding activity of Cra a 4 and in vitro digested samples was reduced by about 50%.The unfoldα-helix and higher surface hydrophobicity of glycated Cra a 4 were identified as the reason for hypo-immunoreactivity.The higher frequency of lysine(13%)in the primary structure and exposure on the surface of the spatial structure was attributed to the hypo-immunoreactivity of Cra a 4 after Maillard reaction.It comprehensively evaluated the effects of Maillard reaction on Cra a 4 and provids new methods for the industrial production of hypoallergenic oyster.The modified Cra a 4 by Maillard reaction could be a hypoallergenic therapeutic agent for allergy to oyster.

Cellular and Molecular Atlas of Peripheral Blood Mononuclear Cells from a Pregnant Woman After Recovery from COVID-19

Objective:This study aimed to investigate the immune response of a pregnant woman who recovered from the coronavirus disease 2019(COVID_(R)S)by using single-cell transcriptomic profiling of peripheral blood mononuclear cells(PBMCs)and to analyze the properties of different immune cell subsets.Methods:PBMCs were collected from the COVID_(R)S patient at 28 weeks of gestation,before a cesarean section.The PBMCs were then analyzed using single-cell RNA sequencing.The transcriptional profiles of myeloid,T,and natural killer(NK)cell subsets were systematically analyzed and compared with those of healthy pregnant controls from a published single-cell RNA sequencing data set.Results:We identified major cell types such as T cells,B cells,NK cells,and myeloid cells in the PBMCs of our COVID_(R)S patient.The increase of myeloid and B cells and decrease of T cells and NK cells in the PBMCs in this patient were quite distinct compared with that in the control subjects.After reclustering and Augur analysis,we found that CD16 monocytes and mucosal-associated invariant T(MAIT)cells were mostly affected within different myeloid,T,and NK cell subtypes in our COVID_(R)S patient.The proportion of CD16 monocytes in the total myeloid population was increased,and the frequency of MAIT cells in the total T and NK cells was significantly decreased in the COVID-RS patient.We also observed significant enrichment of gene sets related to antigen processing and presentation,T-cell activation,T-cell differentiation,and tumor necrosis factor superfamily cytokine production in CD16 monocytes,and enrichment of gene sets related to antigen processing and presentation,response to type II interferon,and response to virus in MAIT cells.Conclusion:Our study provides a single-cell resolution atlas of the immune gene expression patterns in PBMCs from a COVID_(R)S patient.Our findings suggest that CD16-positive monocytes and MAIT cells likely play crucial roles in the maternal immune response against severe acute respiratory syndrome coronavirus 2 infection.These results contribute to a better understanding of the maternal immune response to severe acute respiratory syndrome coronavirus 2 infection and may have implications for the development of effective treatments and preventive strategies for the coronavirus disease 2019 in pregnant women.

Risk Factors and Pregnancy Outcome in Women with a History of Cesarean Section Complicated by Placenta Accreta

Objective:To explore the risk factors and pregnancy outcomes in women with a history of cesarean section complicated by placenta accreta(PA).Methods:This case-control study included clinical data from singleton mothers with a history of cesarean section in 11 public tertiary hospitals in seven provinces of China between January 2017 and December 2017.According to the intraoperative findings after delivery,the study population was divided into PA and non-PA groups.We compared the pregnancy outcomes between the two groups,used multivariate logistic regression to analyze the risk factors for placental accreta.Results:For this study we included 11,074 pregnant women with a history of cesarean section;and of these,869 cases were in the PA group and 10,205 cases were in the non-PA group.Compared with the non-PA group,the probability of postpartum hemorrhage(236/10,205,2.31%vs.283/869,32.57%),severe postpartum hemorrhage(89/10,205,0.87%vs.186/869,21.75%),diffuse intravascular coagulation(3/10,205,0.03%vs.4/869,0.46%),puerperal infection(33/10,205,0.32%vs.12/869,1.38%),intraoperative bladder injury(1/10,205,0.01%vs.16/869,1.84%),hysterectomy(130/10,205,1.27%vs.59/869,6.79%),and blood transfusion(328/10,205,3.21%vs.231/869,26.58%)was significantly increased in the PA group(P<0.05).At the same time,the neonatal birth weight 3250.00(2950.00–3520.00)g vs.2920.00(2530.00–3250.00)g),the probability of neonatal comorbidities(245/10,205,2.40%vs.61/869,7.02%),and the rate of neonatal intensive care unit admission(817/10,205,8.01%vs.210/869,24.17%)also increased significantly(P<0.05).Weight(odds ratio)(OR)=1.03,95%confidence interval(CI):1.01–1.05)),parity(OR=1.18,95%CI:1.03–1.34),number of miscarriages(OR=1.31,95%CI:1.17–1.47),number of previous cesarean sections(OR=2.57,95%CI:2.02–3.26),history of premature rupture of membrane(OR=1.61,95%CI:1.32–1.96),previous cesarean-section transverse incisions(OR=1.38,95%CI:1.12–1.69),history of placenta previa(OR=2.44,95%CI:1.50–3.96),and the combination of prenatal hemorrhage(OR=9.95,95%CI:8.42–11.75)and placenta previa(OR=91.74,95%CI:74.11–113.56)were all independent risk factors for PA.Conclusion:There was an increased risk of adverse outcomes in pregnancies complicated by PA in women with a history of cesarean section,and this required close clinical attention.Weight before pregnancy,parity,number of miscarriages,number of previous cesarean sections,history of premature rupture of membranes,past transverse incisions in cesarean sections,a history of placenta previa,prenatal hemorrhage,and placenta previa were independent risk factors for pregnancies complicated with PA in women with a history of cesarean section.These independent risk factors showed a high value in predicting the risk for placentab accreta in pregnancies of women with a history of cesarean section.

Association of an Increased Risk of Pre-eclampsia and Fetal Growth Restriction in Singleton and Twin Pregnancies with Female Fetuses

Objective: To investigate whether the fetal gender affects the incidence of pre-eclampsia (PE) and fetal growth restriction (FGR) in singleton and twin pregnancies.Methods: This was a 10-year single-center, retrospective, cohort study from January 2009 to January 2019. A total of 57,129 singleton and 3699 twin pregnancies aged between 18-55 years old were recruited at the Third Affiliated Hospital of Guangzhou Medical University, China. We used multivariable logistic regression to analyze the effect of fetal gender on the incidence of PE and FGR.Results: In singleton pregnancies, the incidence rates of PE and FGR with a female fetus were higher than those with a male fetus (6.4% (1713/26,793)vs. 5.9% (1803/30,336),P < 0.05 and 3.5% (932/26,793)vs. 2.4% (745/30,336),P < 0.05, respectively). A female fetus was an independent risk factor for either PE or FGR (adjusted odds ratio: 1.169 or 1.563;95% confidence interval: 1.036-1.319 or 1.349-1.810, respectively). In twin pregnancies, the incidence of early-onset PE was greater in pregnancies with two females compared with two males or one male plus one female (4.6% (46/1003)vs. 4.1% (54/1305)vs. 2.4% (33/1391),P < 0.05). Female-female twins was an independent risk factor for PE (adjusted odds ratio: 1.367, 95% confidence interval: 1.011-1.849), especially early-onset PE.Conclusion: The female fetus was associated with PE in both singleton and twin pregnancies and was also a risk factor of FGR in singleton pregnancies.

Synaptic connections between different sensory channels in CNS of the leech modulate the input of mechanosensory information

The mechanism for transmission of sensory information concerning a specific sensory modality or submodality can be called a sensory channel, including the receptors, sensory pathways and the parts of the central nervous system that further process the sensory information . A certain sensory channel can be activated by a suitable stimulus. Sensory information is transduced by sensory receptor and is further transferred into central nervous system. The characteristics of a modality or submodality are mainly determined by the properties of the corresponding sensory receptors. However, besides the receptor properties, are there any other factors which have influence on the properties of a sensory modality? Is there a kind of gating mechanism existing which could selectively control the inputs of sensory information based on different sensory channels? Here we are trying to answer the above questions by studying the functional relationships between different mechanosensory modalities of leech.