教学课程资源库 更多>>

关键词： Random forest classifier   TCR-pMHC complexes   Binding affinity  

摘要： The activation of T cells is triggered by the interactions of T cell receptors (TCRs) with their epitopes, which are peptides presented by major histocompatibility complex (MHC) on the surfaces of antigen presenting cells (APC). While each TCR can only recognize a specific subset from a large repertoire of peptide-MHC (pMHC) complexes, it is very often that peptides in this subset share little sequence similarity. This is known as the specificity and cross-reactivity of T cells, respectively. The binding affinities between different types of TCRs and pMHC are the major driving force to shape this specificity and cross-reactivity in T cell recognition. The binding affinities, furthermore, are determined by the sequence and structural properties at the interfaces between TCRs and pMHC. Fortunately, a wealth of data on binding and structures of TCR-pMHC interactions becomes publicly accessible in online resources, which offers us the opportunity to develop a random forest classifier for predicting the binding affinities between TCR and pMHC based on the structure of their complexes. Specifically, the structure and sequence of a given complex were projected onto a high-dimensional feature space as the input of the classifier, which was then trained by a large-scale benchmark dataset. Based on the cross-validation results, we found that our machine learning model can predict if the binding affinity of a given TCR-pMHC complex is stronger or weaker than a predefined threshold with an overall accuracy approximately around 75 %. The significance of our prediction was estimated by statistical analysis. Moreover, more than 60 % of binding affinities in the ATLAS database can be successfully classified into groups within the range of 2 kcal/mol. Additionally, we show that TCR-pMHC complexes with strong binding affinity prefer hydrophobic interactions between amino acids with large aromatic rings instead of electrostatic interactions. Our results therefore provide insights to d

关键词： machine learning   multi-label   neural network  

摘要： Multi-label classification algorithms deal with classification problems where a single datapoint can be classified (or labelled) with more than one class (or label) at the same time. Early multi-label approaches like binary relevance consider each label individually and train individual binary classifier models for each label. State-of-the-art algorithms like RAkEL, classifier chains, calibrated label ranking, IBLR-ML+, and BPMLL also consider the associations between labels for improved performance. Like most machine learning algorithms, however, these approaches require careful hyper-parameter tuning, a computationally expensive optimisation problem. There is a scarcity of multi-label classification algorithms that require minimal hyper-parameter tuning. This paper addresses this gap in the literature by proposing CascadeML, a multi-label classification method based on the existing cascaded neural network architecture, which also takes label associations into consideration. CascadeML grows a neural network architecture incrementally (deep as well as wide) in a two-phase process as it learns network weights using an adaptive first-order gradient descent algorithm. This omits the requirement of preselecting the number of hidden layers, nodes, activation functions, and learning rate. The performance of the CascadeML algorithm was evaluated using 13 multi-label datasets and compared with nine existing multi-label algorithms. The results show that CascadeML achieved the best average rank over the datasets, performed better than BPMLL (one of the earliest well known multi-label specific neural network algorithms), and was similar to the state-of-the-art classifier chains and RAkEL algorithms.

关键词： Mediastinum   Lymphoma   Female   Radiotherapy   Thymus  

摘要： Primary mediastinal large B-cell lymphoma (PMBCL) is an aggressive B-cell lymphoma arising from thymic B-cells having clinicopathologic features distinct from systemic diffuse large B-cell lymphoma (DLBCL). PMBCL comprises 2% to 4% of all non-Hodgkin lymphomas (NHL), 7% of DLBCL and seen predominantly in young females with a median age of 35 years at diagnosis. The annual incidence of PMBCL is 0.4 per million with a 5-year survival rate exceeding 70% with improving supportive care and genetic characterization of the disease. Pathogenesis involves dysregulation of Janus kinase-signal transducer and activator of transcription (JAK-STAT), nuclear factor-kB (NF-kB) pathways and amplification of the 9p24.1 region of chromosome 9. PMBCL patients have a prolonged life expectancy necessitating the need for treatment approaches that are based on maximizing cure with minimal long-term toxicity. Due to rarity and its recognition as a distinct entity, therapeutic decisions are guided by clinical presentation, clinician and center experience, and analysis of patients with PMBCL within DLBCL registries. Historically R-CHOP has been the usual first line treatment for PMBCL followed by involved site radiotherapy (ISRT), however clinical practice varies across centers with emerging consensus to avoid upfront RT by utilizing dose intense regimens (DA-EPOCH-R) in younger and fit patients. Prognosis of relapsed refractory PMBCL not responding to salvage chemotherapy is dismal, however there are many emerging options including Brentuximab Vedotin, immune check point inhibitors and chimeric antigen receptor T-cell therapy. In this article, we focus on the pathogenesis, current and evolving treatments, and provide recommendations for optimal management of patients with PMBCL. (C) 2021 Elsevier Inc. All rights reserved.

关键词： Nuclear power plant   Nuclear accidents identification   Deep neural network   Deep learning   Anomaly detection   Time series classification  

摘要： Nuclear power plants (NPPs) are complex systems that are monitored by a team of highly trained operators, that in case of an anomalous event on the NPP, such as an accident, must quickly analyze a high number of variables in order to identify the event. However, due to the great volume of information that must be rapidly analyzed in those situations, misidentifications can occur. Therefore, this study proposes the development of a modular structured, deep learning (DL)-based, NPP accident identification system including anomaly detection and "don't know" response capability. In order to assess the system capacity, experiments were conducted using a realistic study case using 16 operational situations of main operational Brazilian PWR NPP. As a result, the best configuration of DL models for the system achieved an average accuracy of 100% in terms of correct "don't know" identification, while still retaining the accuracy of 99.82% on correct event identifications. (c) 2021 Elsevier Ltd. All rights reserved.

摘要： This case presents Primary Squamous Cell Carcinoma of The Endometrium with cytological and immunocytological characteristics in specimens from endometrial aspirate, Pap smear and peritoneal washing. Initial TVUS and clinical examination showed no abnormalities, and the first curettage sample was interpreted as benign while the second was histopathologically inconclusive. The final histopathology diagnosis in the hysterectomy specimen confirmed the diagnosis of this rare tumour.

关键词： Cancer models   Cell signalling   Medicine/Public Health   general   Internal Medicine   Intensive / Critical Care Medicine   Cancer Research   Oncology   Hematology  

摘要： Pediatric myelodysplastic syndromes (MDS) are a heterogeneous disease group associated with impaired hematopoiesis, bone marrow hypocellularity, and frequently have deletions involving chromosome 7 (monosomy 7). We and others recently identified heterozygous germline mutations in SAMD9 and SAMD9L in children with monosomy 7 and MDS. We previously demonstrated an antiproliferative effect of these gene products in non-hematopoietic cells, which was exacerbated by their patient-associated mutations. Here, we used a lentiviral overexpression approach to assess the functional impact and underlying cellular processes of wild-type and mutant SAMD9 or SAMD9L in primary mouse or human hematopoietic stem and progenitor cells (HSPC). Using a combination of protein interactome analyses, transcriptional profiling, and functional validation, we show that SAMD9 and SAMD9L are multifunctional proteins that cause profound alterations in cell cycle, cell proliferation, and protein translation in HSPCs. Importantly, our molecular and functional studies also demonstrated that expression of these genes and their mutations leads to a cellular environment that promotes DNA damage repair defects and ultimately apoptosis in hematopoietic cells. This study provides novel functional insights into SAMD9 and SAMD9L and how their mutations can potentially alter hematopoietic function and lead to bone marrow hypocellularity, a hallmark of pediatric MDS.

关键词： Merkel cell carcinoma   Merkel cell carcinoma survival   tumor primary site   tumor site   unknown primary site  

摘要： Background: Merkel cell carcinoma (MCC) primary site has not been fully investigated as a potential prognostic factor. Objective: To determine the incidence by tumor primary site of death due to MCC. Methods: We undertook a retrospective analysis of the Survival, Epidemiology, and End Results database. MCC patients treated between 1973 and 2016 were grouped by tumor primary site and a competing risks analysis was performed to test the impact of primary site on disease-specific death. Cumulative incidence of Merkel cell carcinoma-specific mortality (CMMI) at 5 years was estimated for each primary site. Results: Of 9407 MCC patients identified, 6305 (67.0%) had localized disease, 2397 (25.5%) had regional metastasis, and 705 (7.5%) had distant metastasis. Tumor primary site was predictive of CMMI and varied by stage at diagnosis. Tumors involving the scalp/neck carried the highest CMMI among localized MCC (26.0%). Tumors involving the lip had the highest CMMI among MCC with regional metastasis (56.7%) and distant metastasis (82.1%). Limitations: Tumor size data were missing for a large proportion of patients, precluding stratification by stage according to current American Joint Committee on Cancer guidelines. Conclusions: Probability of MCC disease-specific death varies by primary site. The primary site of the tumor may be useful as a prognostic indicator for MCC. ( J Am Acad Dermatol 2021;85:125966.)

关键词： Calcium deposition   Endometrial epithelial cell   Laying hen   Zinc transporter   Zn-methionine  

摘要： For birds, the uterus is an important part for eggshell mineralization, and the establishment of the endometrial epithelial cell (EEC) model was beneficial to the study of uterine function. This study was conducted to establish a culture model of primary EECs of laying hens and explore the effects of zinc on primary EEC proliferation, zinc transport, and calcium deposition in vitro. The EECs were isolated and cultured via type I collagenase digestion, and in the logarithmic phase during 2-5 days, and then reached the plateau phase on the 7th day of inoculation. Results showed that the proliferation of EECs treated by 50 mu M ZnSO4 or zinc-methionine (Zn-Met) were markedly promoted at 24-h or 48-h treating time (P < 0.05). In later experiments, the EECs were divided into three groups, involving a control group (no zinc treated), ZnSO4 group (50 mu M zinc treated) and a Zn-Met group (50 mu M zinc treated). Results showed the relative fluorescence intensity of Ca2+ in the Zn-Met group was significantly higher than that in the control group (P < 0.05). As for zinc transporters, it was only observed that mRNA levels of metallothionein (MT) in EECs showed a significant difference (P < 0.05) between the Zn-Met group and the control. In conclusion, the EECs of laying hens isolated by scraping and digested collagenase I were with better adherent growth. Moreover, Zn-Met can increase intracellular Ca2+ concentration and upregulate expressions of MT mRNA in the EECs of laying hens.

关键词： Blood–brain barrier   Central nervous system infections   Biomedicine   general   Neurosciences   Behavioral Sciences   Biological Techniques   Neurobiology   Animal Genetics and Genomics  

摘要： Coronavirus disease 2019 (COVID-19) can damage cerebral small vessels and cause neurological symptoms. Here we describe structural changes in cerebral small vessels of patients with COVID-19 and elucidate potential mechanisms underlying the vascular pathology. In brains of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2)-infected individuals and animal models, we found an increased number of empty basement membrane tubes, so-called string vessels representing remnants of lost capillaries. We obtained evidence that brain endothelial cells are infected and that the main protease of SARS-CoV-2 (M-pro) cleaves NEMO, the essential modulator of nuclear factor-kappa B. By ablating NEMO, M-pro induces the death of human brain endothelial cells and the occurrence of string vessels in mice. Deletion of receptor-interacting protein kinase (RIPK) 3, a mediator of regulated cell death, blocks the vessel rarefaction and disruption of the blood-brain barrier due to NEMO ablation. Importantly, a pharmacological inhibitor of RIPK signaling prevented the M-pro-induced microvascular pathology. Our data suggest RIPK as a potential therapeutic target to treat the neuropathology of COVID-19. A novel study led by scientists in Lubeck, Germany, shows that SARS-CoV-2-infected brain endothelial cells undergo cell death due to the cleavage of NEMO by the viral protease M-pro, potentially causing cerebral COVID-19 and 'long COVID' symptoms.