Unsupervised Set Research Articles

12408 Efnb2 Controls Pituitary Gland Development By Regulating Proliferation Of Pituitary Stem Cells

Abstract Disclosure: A. Gualtieri: None. L. Gomez-Corral: None. J. Nicholson: None. R. Tan: None. L. Guasti: None. C. Gaston-Massuet: None. Efnb2 encodes for the ligand Ephrinb2 that binds to its cognate Eph receptor, which plays an integral role in angiogenesis, stem cell regulation and tumorigenesis. Using a pituitary-specific Cre-driver (Hesx1Cre), we conditionally deleted Efnb2 from the early pituitary stem/progenitor cells (PSCs) of the developing pituitary gland. We found that Efnb2 is expressed in PSCs both during embryogenesis and adulthood, suggesting its involvement in PSCs maintenance. Here we show that genetic ablation of Efnb2 (Efnb2fl/fl;Hesx1Cre/+) results in severe pituitary hyperplasia, abnormal gland morphogenesis and delay in the terminal differentiation of hormone-producing cells. In order to understand the underlying molecular mechanisms by which Efnb2 deletion leads to pituitary abnormalities during embryogenesis, we performed transcriptomic analyses using mRNA-Seq at early stages of pituitary development on Efnb2+/+ and Efnb2−/− cell populations. Interestingly, unsupervised gene set enrichment analyses of the transcriptomic results identified Efnb2 as a key negative regulator of PSCs proliferation, whilst involved in the regulation of epithelial integrity and epithelial to mesenchymal transition (EMT). Functional assays performed to validate transcriptomic analyses revealed that Efnb2-null PSCs hyper-proliferate both in vitro and in vivo and downregulate the expression of genes involved in EMT, indicating a role of Efnb2 in these two developmental pathways. Indeed, whilst Efnb2-null PSCs within the marginal zone hyper-proliferate, they downregulate genes of epithelial integrity, resulting in an abnormal EMT transcriptomic profile. Additionally, these events lead to a significant delay of the pituitary cell lineage commitment program, leading to severe reduction in the expression of Pit1, Pomc1 and Gsu commitment markers. Taken together, our results reveal a novel role for Efnb2 (Ephrinb2) as key regulator of pituitary gland development. Efnb2 negatively regulates the proliferation of pituitary stem/progenitor cells controlling pituitary organ size while controlling the epithelial integrity of the stem cell niche. Presentation: 6/2/2024

Exploring the transcriptomic and metabolomic profiles of adipose tissues: Insights and implications for fat grafting

BackgroundExpanding on our prior research on murine fat grafts' metabolic shift, this study delves deeper into the metabolic profiles of human adipose tissues, specifically the superficial subcutaneous adipose tissue (SSAT) and deep subcutaneous adipose tissue (DSAT). MethodsUtilizing RNA-sequencing, metabolome, and metabolic flux analyses, we examined SSAT and DSAT samples obtained during deep inferior epigastric perforator flap breast reconstructions. Transcript data underwent unsupervised hierarchical clustering and Gene set enrichment analysis. Metabolomics involved analyzing samples for cationic and anionic metabolites via Capillary Electrophoresis Time-of-Flight Mass Spectrometry (CE-TOFMS), followed by principal component analysis (PCA) and heat map generation. Additionally, primary adipocytes from SSAT and DSAT were assessed using the Seahorse® extracellular flux analyzer. ResultsPCA and heat map analyses highlighted distinct transcriptome and metabolome differences between SSAT and DSAT. SSAT predominantly upregulated genes linked to adipogenesis (false discovery rate [FDR] q < 0.0001), oxidative phosphorylation (FDR q < 0.0001), fatty acid metabolism (FDR q <0.0001), and glycolysis (FDR q = 0.001). In contrast, DSAT showed a significant upregulation in inflammatory response genes (FDR q <0.05). Metabolite analysis revealed an abundance of glycolytic metabolites in SSAT, whereas DSAT was rich in metabolites associated with fatty acid metabolism and oxidative phosphorylation. Cellular flux analysis further confirmed SSAT's elevated glycolysis and spare oxidative phosphorylation capacities. ConclusionOur results highlight the metabolic uniqueness of SSAT and DSAT in humans, with SSAT exhibiting superior metabolic flexibility. The implications of these metabolic differences, especially in fat grafting procedures, necessitate further research and exploration in future studies. Data availabilityThe ribonucleic acid sequencing data were deposited in the Gene Expression Omnibus database (GSE216803). https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE216803 (GEO token: qnsngiuqplqdpaf).

AlpaPICO: Extraction of PICO frames from clinical trial documents using LLMs

In recent years, there has been a surge in the publication of clinical trial reports, making it challenging to conduct systematic reviews. Automatically extracting Population, Intervention, Comparator, and Outcome (PICO) from clinical trial studies can alleviate the traditionally time-consuming process of manually scrutinizing systematic reviews. Existing approaches of PICO frame extraction involves supervised approach that relies on the existence of manually annotated data points in the form of BIO label tagging. Recent approaches, such as In-Context Learning (ICL), which has been shown to be effective for a number of downstream NLP tasks, require the use of labeled examples. In this work, we adopt ICL strategy by employing the pretrained knowledge of Large Language Models (LLMs), gathered during the pretraining phase of an LLM, to automatically extract the PICO-related terminologies from clinical trial documents in unsupervised set up to bypass the availability of large number of annotated data instances. Additionally, to showcase the highest effectiveness of LLM in oracle scenario where large number of annotated samples are available, we adopt the instruction tuning strategy by employing Low Rank Adaptation (LORA) to conduct the training of gigantic model in low resource environment for the PICO frame extraction task. More specifically, both of the proposed frameworks utilize AlpaCare as base LLM which employs both few-shot in-context learning and instruction tuning techniques to extract PICO-related terms from the clinical trial reports. We applied these approaches to the widely used coarse-grained datasets such as EBM-NLP, EBM-COMET and fine-grained datasets such as EBM-NLPrev and EBM-NLPh. Our empirical results show that our proposed ICL-based framework produces comparable results on all the version of EBM-NLP datasets and the proposed instruction tuned version of our framework produces state-of-the-art results on all the different EBM-NLP datasets. Our project is available at https://github.com/shrimonmuke0202/AlpaPICO.git.

A GAN-based anomaly detector using multi-feature fusion and selection

In numerous applications, abnormal samples are hard to collect, limiting the use of well-established supervised learning methods. GAN-based models which trained in an unsupervised and single feature set manner have been proposed by simultaneously considering the reconstruction error and the latent space deviation between normal samples and abnormal samples. However, the ability to capture the input distribution of each feature set is limited. Hence, we propose an unsupervised and multi-feature model, Wave-GANomaly, trained only on normal samples to learn the distribution of these normal samples. The model predicts whether a given sample is normal or not by its deviation from the distribution of normal samples. Wave-GANomaly fuses and selects from the wave-based features extracted by the WaveBlock module and the convolution-based features. The WaveBlock has proven to efficiently improve the performance on image classification, object detection, and segmentation tasks. As a result, Wave-GANomaly achieves the best average area under the curve (AUC) on the Canadian Institute for Advanced Research (CIFAR)-10 dataset (94.3%) and on the Modified National Institute of Standards and Technology (MNIST) dataset (91.0%) when compared to existing state-of-the-art anomaly detectors such as GANomaly, Skip-GANomaly, and the skip-attention generative adversarial network (SAGAN). We further verify our method by the self-curated real-world dataset, the result show that our method is better than GANomaly which only use single feature set for training the model.

DSLSM: Dual-kernel-induced statistic level set model for image segmentation

Even though the level set method is driving progress in image segmentation, its performance is still affected adversely in the presence of severe intensity inhomogeneity and multi-noise. Therefore, developing an accurate, robust, and fast technique to solve this becomes a pressing need. In this paper, we propose DSLSM, a novel unsupervised statistic level set model based on dual-kernel-induced local similarity measures, which not only can effectively handle severe intensity inhomogeneity and multi-noise, but also stabilizes and accelerates the evolution process. Specifically, we first define a novel dual-kernel-induced data term, which stacks multiple kernel functions and manipulates local similarity measurement via the Jensen–Shannon divergence, aiming at extracting more discriminative higher-dimension features and correlations between adjacent pixels. Then, a novel rectified morphological scheme is proposed to numerically implement our method, which not only precludes the issues of numerical instability and level set function degradation but significantly accelerates the contour evolution. In addition, an inverted pyramid-style scheduling strategy is proposed to explore the complementary utilization of local and non-local contextual features at different evolution phases to avoid tedious heuristics parameter tuning for the size of local regions. Extensive experiments on multiple challenging segmentation benchmarks demonstrate the effectiveness and superiority of our DSLSM over recent state-of-the-art methods.

Balanced and robust unsupervised Open Set Domain Adaptation via joint adversarial alignment and unknown class isolation

Open Set Domain Adaptation aims to implement knowledge transfer from the label-rich source domain to the label-sparse or unlabeled target domain in the scenario where the target domain contains unknown classes that are absent in the source domain. Most prevalent approaches produce pseudo labels or weights to separate the known and unknown classes in the target domain. Subsequently, these methods mitigate the domain discrepancy among known classes through adversarial alignment or by minimizing metrics that quantify the inter-domain dissimilarity. Nevertheless, the pseudo labels and weights generated by these methods can significantly deviate from the ground truth in the presence of substantial domain discrepancy, resulting in diminished accuracy when detecting unknown classes. To address this issue, we propose a balanced and robust unsupervised open set domain adaptation method which follows an end-to-end manner. Specifically, we initially extract new feature representations by a feature extractor. Subsequently, we establish the Ensemble-Binary Deviation Network to compute the deviation scores of samples in the new feature space. We maximize the divergence in deviation scores among various classes to sharpen the inter-class boundaries. Simultaneously, we employ two classifiers in the new feature space to align the feature distributions. Particularly, we initially identify the hard-to-classify samples by maximizing the discrepancy of two classifiers. Secondly, we enforce the feature extractor to minimize the discrepancy of two classifiers, aligning the feature distribution and extracting more discriminative new features. Extensive experiments on three benchmarks verify that our method achieves a balanced and robust performance, manifesting an enhancement ranging from approximately 4.9% to 7.0% over the compared methods.

Bridging the Theoretical Bound and Deep Algorithms for Open Set Domain Adaptation.

In the unsupervised open set domain adaptation (UOSDA), the target domain contains unknown classes that are not observed in the source domain. Researchers in this area aim to train a classifier to accurately: 1) recognize unknown target data (data with unknown classes) and 2) classify other target data. To achieve this aim, a previous study has proven an upper bound of the target-domain risk, and the open set difference, as an important term in the upper bound, is used to measure the risk on unknown target data. By minimizing the upper bound, a shallow classifier can be trained to achieve the aim. However, if the classifier is very flexible [e.g., deep neural networks (DNNs)], the open set difference will converge to a negative value when minimizing the upper bound, which causes an issue where most target data are recognized as unknown data. To address this issue, we propose a new upper bound of target-domain risk for UOSDA, which includes four terms: source-domain risk, ϵ -open set difference ( ∆ϵ ), distributional discrepancy between domains, and a constant. Compared with the open set difference, ∆ϵ is more robust against the issue when it is being minimized, and thus we are able to use very flexible classifiers (i.e., DNNs). Then, we propose a new principle-guided deep UOSDA method that trains DNNs via minimizing the new upper bound. Specifically, source-domain risk and ∆ϵ are minimized by gradient descent, and the distributional discrepancy is minimized via a novel open set conditional adversarial training strategy. Finally, compared with the existing shallow and deep UOSDA methods, our method shows the state-of-the-art performance on several benchmark datasets, including digit recognition [modified National Institute of Standards and Technology database (MNIST), the Street View House Number (SVHN), U.S. Postal Service (USPS)], object recognition (Office-31, Office-Home), and face recognition [pose, illumination, and expression (PIE)].

Unsupervised Category-Specific Partial Point Set Registration via Joint Shape Completion and Registration.

We propose a self-supervised method for partial point set registration. Although recently proposed learning-based methods demonstrate impressive registration performance on full shape observations, these methods often suffer from performance degradation when dealing with partial shapes. To bridge the performance gap between partial and full point set registration, we propose to incorporate a shape completion network to benefit the registration process. To achieve this, we introduce a learnable latent code for each pair of shapes, which can be regarded as the geometric encoding of the target shape. By doing so, our model does not require an explicit feature embedding network to learn the feature encodings. More importantly, both our shape completion and point set registration networks take the shared latent codes as input, which are optimized simultaneously with the parameters of two decoder networks in the training process. Therefore, the point set registration process can benefit from the joint optimization process of latent codes, which are enforced to represent the information of full shapes instead of partial ones. In the inference stage, we fix the network parameters and optimize the latent codes to obtain the optimal shape completion and registration results. Our proposed method is purely unsupervised and does not require ground truth supervision. Experiments on the ModelNet40 dataset demonstrate the effectiveness of our model for partial point set registration.

Panomics reveals patient individuality as the major driver of colorectal cancer progression

BackgroundColorectal cancer (CRC) is one of the most prevalent cancers, with over one million new cases per year. Overall, prognosis of CRC largely depends on the disease stage and metastatic status. As precision oncology for patients with CRC continues to improve, this study aimed to integrate genomic, transcriptomic, and proteomic analyses to identify significant differences in expression during CRC progression using a unique set of paired patient samples while considering tumour heterogeneity.MethodsWe analysed fresh-frozen tissue samples prepared under strict cryogenic conditions of matched healthy colon mucosa, colorectal carcinoma, and liver metastasis from the same patients. Somatic mutations of known cancer-related genes were analysed using Illumina's TruSeq Amplicon Cancer Panel; the transcriptome was assessed comprehensively using Clariom D microarrays. The global proteome was evaluated by liquid chromatography-coupled mass spectrometry (LC‒MS/MS) and validated by two-dimensional difference in-gel electrophoresis. Subsequent unsupervised principal component clustering, statistical comparisons, and gene set enrichment analyses were calculated based on differential expression results.ResultsAlthough panomics revealed low RNA and protein expression of CA1, CLCA1, MATN2, AHCYL2, and FCGBP in malignant tissues compared to healthy colon mucosa, no differentially expressed RNA or protein targets were detected between tumour and metastatic tissues. Subsequent intra-patient comparisons revealed highly specific expression differences (e.g., SRSF3, OLFM4, and CEACAM5) associated with patient-specific transcriptomes and proteomes.ConclusionOur research results highlight the importance of inter- and intra-tumour heterogeneity as well as individual, patient-paired evaluations for clinical studies. In addition to changes among groups reflecting CRC progression, we identified significant expression differences between normal colon mucosa, primary tumour, and liver metastasis samples from individuals, which might accelerate implementation of precision oncology in the future.

Automatic Classification of Galaxy Morphology: A Rotationally-invariant Supervised Machine-learning Method Based on the Unsupervised Machine-learning Data Set

Classification of galaxy morphology is a challenging but meaningful task for the enormous amount of data produced by the next-generation telescope. By introducing the adaptive polar-coordinate transformation, we develop a rotationally-invariant supervised machine-learning (SML) method that ensures consistent classifications when rotating galaxy images, which is always required to be satisfied physically, but difficult to achieve algorithmically. The adaptive polar-coordinate transformation, compared with the conventional method of data augmentation by including additional rotated images in the training set, is proved to be an effective and efficient method in improving the robustness of the SML methods. In the previous work, we generated a catalog of galaxies with well-classified morphologies via our developed unsupervised machine-learning (UML) method. By using this UML data set as the training set, we apply the new method to classify galaxies into five categories (unclassifiable, irregulars, late-type disks, early-type disks, and spheroids). In general, the result of our morphological classifications following the sequence from irregulars to spheroids agrees well with the expected trends of other galaxy properties, including Sérsic indices, effective radii, nonparametric statistics, and colors. Thus, we demonstrate that the rotationally-invariant SML method, together with the previously developed UML method, completes the entire task of automatic classification of galaxy morphology.

Outlier Detection in Test Samples using Standard Deviation and Unsupervised Training Set Selection

Outlier detection is a technique to identify and remove significantly different data from the more correct and consistent data in a data set. Outlier data can have negative impact on classification and clustering performance; that should be identified and removed to improve the classification efficiency. Regardless of whether a classifying technique classifies an outlier correctly, the very notion of identifying a data as outlier is of great significance. In this paper, a new approach is proposed for outlier data detection within a test data set along with unsupervised training set selection. The selected training set is used for two-step classification. After unsupervised clustering the training set, the closest cluster to a test sample is selected using the Euclidean distance measure. Then, the outlier in the test sample is identified with the concepts of standard deviation and mean value. The results showed by evaluating the distance of each sample of the test set with the new selected data set. the accuracy of the classifiers is enhanced after detection and elimination of outlier data.

Machine Learning Pipeline with Feature Engineering Provides Robust Diagnostic Predictions in Chronic Lymphocytic Leukemia, Accelerated and Transformed Phases

BackgroundChronic lymphocytic leukemia/small lymphocytic lymphoma (CLL) is an indolent disease. However, a small subset of cases may progress to accelerated CLL (aCLL) and eventually transform to diffuse large B-cell lymphoma, also known as Richter transformation (RT). Core-needle biopsies of CLL in accelerated and transformed phases present a plethora of diagnostic challenges, hindering a confident and precise morphologic assessment. To overcome these impediments, we propose a high-throughput diagnostic pipeline empowered by deep learning to discover and characterize intrinsic cell populations and help boost diagnostic accuracy.Material & MethodsWe collected 193 biopsies from 125 patients, including 69 CLL slides from 44 patients, 44 aCLL slides from 34 patients, and 80 RT slides from 47 patients. Our computational pipeline contained the following 7 steps (Figure 1): 1) ROI selection, 2) stain normalization ; 3) cell segmentation through transfer learning, using a pre-trained deep learning model (HoVer-Net); 4) quality control of automated cell segmentation; 5) profiling cell populations from three different perspectives, including grouping of cells into large or small subtypes using supervised learning, discovering the intrinsic cell subpopulations with unsupervised learning, and mixing cells for indiscriminate profiling; 6) pruning uninformative features by quantifying feature importance via impurity analysis; 7) systematically evaluating the diagnostic performance of the three cellular profiling methods as well as feature fusion and selection, followed by two validation strategies: The first one aimed to stratify patients into training and testing cohorts to balance key clinicopathologic factors (one-shot validation); and the second one aimed to randomly split patients into training or testing sets, followed by repeat splitting for 100 times (repeated cross-validation).ResultsFirst, we sought to define cells into large or small populations, where cell size cutoff was learned to maximize pairwise separation among three disease types (CLL, aCLL, and RT). We then measured large cell ratios, correlations between large and small cell intensity and density, and mean cell to the nearest-neighbor distance, and labeled the extracted attributes as “supervised feature set” (Figure 1E). Second, we applied an unsupervised learning (i.e., spectral clustering) to detect the intrinsic cell subpopulations based on morphology and intensity. Interestingly, three cell phenotypes were uncovered, which we termed as “CLL-like,” “aCLL-like,” and “RT-like” cells. The ratios of the three cell types in each ROI were computed and labeled as “unsupervised feature set” (Figure 1F). Third, we analyzed cells as one cohort, and computed the mean cell size and intensity, cellular density, and cell to its nearest-neighbor distance, this population was labeled as “mixed cell feature set” as a whole (Figure 1G). lastly, we applied feature selection of fused feature sets, where feature importance was calculated via impurity analysis. Subsequently, 6 out of the total 17 features were pruned (Figure 1H). When testing the three feature sets separately, we observed that the “mixed cell feature set” achieved the best performance (AUC=0.951; n=4 features) followed by the “unsupervised feature set” (AUC=0.902; n=3 features) and “supervised feature set” (AUC=0.829; n=10 features). By integrating the three feature sets, we obtained an accuracy of 0.874 and an area under the curve (AUC) of 0.961 in one-shot validation and a mean accuracy of 0.831 and AUC of 0.952 in repeated cross-validation, surpassing the performance obtained by solely adopting a single feature set at a time. Application of feature selection to fused feature sets further boosted the accuracy to 0.883 and AUC to 0.966 via one-shot validation, and mean accuracy of 0.842 and mean AUC of 0.959 via repeated cross-validation (Figure 1I).ConclusionThe “mixed feature set” achieved higher diagnostic accuracy in comparison to the “supervised” and “unsupervised” feature sets, emphasizing the power of characterizing heterogeneous cell populations. The synergy of three feature sets validates the hypothesis that integrating different ways of cellular phenotyping may optimize the predictive power. Eliminating less informative features further enhances diagnostic accuracy, highlighting the importance of adopting meaningful attributes. [Display omitted] DisclosuresKhoury: Stemline Therapeutics: Research Funding; Kiromic: Research Funding; Angle: Research Funding.

Vehicle Dimensions Based Passenger Car Classification using Fuzzy and Non-Fuzzy Clustering Methods

There has been globally continuous growth in passenger car sizes and types over the past few decades. To assess the development of vehicular specifications in this context and to evaluate changes in powertrain technologies depending on surrounding frame conditions, such as charging stations and vehicle taxation policy, we need a detailed understanding of the vehicle fleet composition. This paper aims therefore to introduce a novel mathematical approach to segment passenger vehicles based on dimensions features using a means fuzzy clustering algorithm, Fuzzy C-means (FCM), and a non-fuzzy clustering algorithm, K-means (KM). We analyze the performance of the proposed algorithms and compare them with Swiss expert segmentation. Experiments on the real data sets demonstrate that the FCM classifier has better correlation with the expert segmentation than KM. Furthermore, the outputs from FCM with five clusters show that the proposed algorithm has a superior performance for accurate vehicle categorization because of its capacity to recognize and consolidate dimension attributes from the unsupervised data set. Its performance in categorizing vehicles was promising with an average accuracy rate of 79% and an average positive predictive value of 75%.

A cloud-based condition monitoring system for fault detection in rotating machines using PROFINET process data

This work presents a methodology for a cloud-based condition monitoring system for fault detection and identification in rotating machines, such as uncoupling, angular and parallel misalignment, by data mining PROFINET network and PROFIdrive profile process data. The proposed methodology involves a new strategy for feature selection of unsupervised data set and employs SVM (Support Vector Machine) and OCSVM (One-Class Support Vector Machine) for operation status classification. The present diagnostic system represents a low-cost solution to the manufacturing process of small and medium enterprises, because it does not require dedicated sensors for fault detection and high featured hardware, and it employs an online cloud-based services. The experimental tests resulted in an accuracy between 87.5% and 100%, and high robustness among different operating conditions. In addition, the proposed feature selection strategy reduced the total execution time by 97.5%.

Open Set Domain Adaptation: Theoretical Bound and Algorithm.

The aim of unsupervised domain adaptation is to leverage the knowledge in a labeled (source) domain to improve a model's learning performance with an unlabeled (target) domain-the basic strategy being to mitigate the effects of discrepancies between the two distributions. Most existing algorithms can only handle unsupervised closed set domain adaptation (UCSDA), i.e., where the source and target domains are assumed to share the same label set. In this article, we target a more challenging but realistic setting: unsupervised open set domain adaptation (UOSDA), where the target domain has unknown classes that are not found in the source domain. This is the first study to provide learning bound for open set domain adaptation, which we do by theoretically investigating the risk of the target classifier on unknown classes. The proposed learning bound has a special term, namely, open set difference, which reflects the risk of the target classifier on unknown classes. Furthermore, we present a novel and theoretically guided unsupervised algorithm for open set domain adaptation, called distribution alignment with open difference (DAOD), which is based on regularizing this open set difference bound. The experiments on several benchmark data sets show the superior performance of the proposed UOSDA method compared with the state-of-the-art methods in the literature.

Cross-Dataset Hyperspectral Image Classification Based on Adversarial Domain Adaptation

The cross-data set knowledge is vital for hyperspectral image classification, which can reduce the dependence on the sample quantity by transferring knowledge from other data sets and improve the training efficiency by sharing knowledge between different data sets. However, due to the capturing environment change and imaging equipment difference, domain shift troubles the exploitation of the cross-data set knowledge. To address the aforementioned issue, this article proposes an unsupervised cross-data set hyperspectral image classification method based on adversarial domain adaptation. The proposed method, which employs multiple classifiers to build a discriminator and uses variational autoencoders to constitute a generator, works in an adversarial manner to drive the target samples under the support of the source domain. In particular, the classification error and the classification disagreement are considered in the objective function, which helps to align different domains while keeping the boundaries of different classes. Experimental results of the multidomain data set demonstrate that the proposed method can transfer and share cross-data set knowledge and achieve state-of-the-art performance without using the labeled information of the target data set.

Extracellular glutamate and IDH1R132H inhibitor promote glioma growth by boosting redox potential.

Somatic mutations of the isocitrate dehydrogenase 1 (IDH1) gene, mostly substituting Arg132 with histidine, are associated with better patient survival, but glioma recurrence and progression are nearly inevitable, resulting in disproportionate morbidity and mortality. Our previous studies demonstrated that in contrast to hemizygous IDH1R132H (loss of wild-type allele), heterozygousIDH1R132H is intrinsically glioma suppressive but its suppression of three-dimensional (3D) growth is negated by extracellular glutamate and reducing equivalent. This study sought to understand the importance of 3D culture in IDH1R132H biology and the underlying mechanism of the glutamate effect. RNA sequencing data of IDH1R132H-heterozygous and IDH1R132H-hemizygous glioma cells cultured under two-dimensional (2D) and 3D conditions were subjected to unsupervised hierarchal clustering and gene set enrichment analysis. IDH1R132H-heterozygous and IDH1R132H-hemizygous tumor growth were compared in subcutaneous and intracranial transplantations. Short-hairpin RNA against glutamate dehydrogenase 2 gene(GLUD2) expression was employed to determine the effects of glutamate and the mutant IDH1 inhibitor AGI-5198 on redox potential in IDH1R132H-heterozygous cells. In contrast to IDH1R132H-heterozygous cells, 3D-cultured but not 2D-cultured IDH1R132H-hemizygous cells were clustered with more malignant gliomas, possessed the glioblastoma mesenchymal signature, and exhibited aggressive tumor growth. Although both extracellular glutamate and AGI-5198 stimulated redox potential for 3D growth of IDH1R132H-heterozygous cells, GLUD2 expression was required for glutamate, but not AGI-5198, stimulation. 3D culture is more relevant to IDH1R132H glioma biology. The importance of redox homeostasis in IDH1R132H glioma suggests that metabolic pathway(s) can be explored for therapeutic targeting, whereas IDH1R132H inhibitors may have counterproductive consequences in patient treatment.

Brain tumor detection based on extreme learning

Gliomas are dreadful and common type of brain tumor. Therefore, treatment planning is significant to increase the survival rate of gliomas patients. The large structural and spatial variation between tumors makes an automated detection more challenging. Brain magnetic resonance imaging is utilized for tumor evaluation on the basis of automated segmentation and classification methods. In this work, triangular fuzzy median filtering is applied for image enhancement that helps in accurate segmentation based on unsupervised fuzzy set method. Gabor features are extracted across each candidate’s lesions, and similar texture (ST) features are calculated. These ST features are supplied to extreme learning machine (ELM), and regression ELM leaves one out for tumor classification. The technique is evaluated on BRATS 2012, 2013, 2014 and 2015 challenging datasets as well as on 2013 Leader board. The proposed approach shows better results and less computational time.

Inhibiting EZH2 function prevents and treats chronic graft vs host disease (cGVHD) by disrupting germinal center formation and plasma cell maturation

Abstract To determine whether targeting EZH2, a histone methyltransferase necessary for germinal center (GC) formation, could reduce murine cGVHD (cGVHD), we first used a clinically relevant murine model of cGVHD, which recapitulates multi-organ cGVHD with bronchiolitis obliterans (BOS). Bone marrow (BM) from EZH2 fl/fl CD19 Cre B6 mice and WT T cells, or WT B6 BM and EZH2 fl/fl CD4 Cre B6 T cells were given to conditioned B10.BR mice. EZH2 expression in BM and T cells was obligatory for cGVHD. To determine whether EZH2 function could be targeted to treat cGVHD, we tested a novel EZH2 small molecule inhibitor, JQ5 with a favorable toxicity and efficacy profile to other inhibitors. Mice with established cGVHD given JQ5 had dramatically improved cGVHD. Similar results were seen in a sclerodermatous cGVHD model. To assess possible causes of this improvement we performed RNA-seq and ATAC-seq data of MACS purified CD19+ B cells obtained from control and JQ5 treated BOS cGVHD mice. Unsupervised gene set enrichment analysis and weighted gene correlation network analysis showed that among gene clusters associated with JQ5 treatment there was an enrichment for genes expressed in naïve B cells compared to plasma cells (GSEA4142) and an impairment of genes associated with late stage GC B cells compared to early follicular B cells (GSEA11961). This suggested that JQ5 impaired GC B cell and plasma cell maturation. Consistent with these data, splenic follicular cells, GC size and number, and lung resident PCs were reduced compared to cGVHD controls. Taken together these results show that EZH2 is necessary for the initiation of cGVHD and show for the first time that EZH2 inhibition using a novel inhibitor JQ5 can reverse established cGVHD in mice with BOS or scleroderma.

Designing a Classifier Using Unsupervised Learning and Rough Set Theory

Designing a Classifier Using Unsupervised Learning and Rough Set Theory