Monocytes contribute to immune responses as a source for subsets of dendritic cells and macrophages. Human blood monocytes are classified as classical, non-classical and intermediate cells. However, the particular functions of these subsets have been hard to define, with conflicting results and significant overlaps. One likely reason for these ambiguities is in the heterogeneity of these monocyte subsets regrouping cells with divergent functions. To better define monocyte populations, we have analysed expression of 17 markers by multicolour flow cytometry in samples obtained from 28 control donors. Data acquisition was tailored to detect populations present at low frequencies. Our results reveal the existence of novel monocyte subsets detected as larger CD14+ cells that were CD16+ or CD16neg. These large monocytes differed from regular, smaller monocytes with respect to expression of various cell surface molecules, such as FcR, chemokine receptors, and adhesion molecules. Unsupervised multidimensional analysis confirmed the existence of large monocytes and revealed interindividual variations that were grouped according to unique patterns of expression of adhesion molecules CD62L, CD49d, and CD43. Distinct inflammatory responses to TLR agonists were found in small and large monocytes. Overall, refining the definition of monocyte subsets should lead to the identification of populations with specific functions.An amendment to this paper has been published and can be accessed via a link at the top of the paper.Exosomes are secreted extracellular vesicles with lipid bilayer membranes. They are emerging as a new category of messengers that facilitate cross-talk between cells, tissues, and organs. Thus, a critical demand arises for the development of a sensitive and non-invasive tracking system for endogenous exosomes. We have generated a genetic mouse model that meets this goal. The Nano-luciferase (NanoLuc) reporter was fused with the exosome surface marker CD63 for exosome labeling. The cardiomyocyte-specific αMHC promoter followed by the loxP-STOP-loxP cassette was engineered for temporal and spatial labeling of exosomes originated from cardiomyocytes. The transgenic mouse was bred with a tamoxifen-inducible Cre mouse (Rosa26Cre-ERT2) to achieve inducible expression of CD63NanoLuc reporter. The specific labeling and tissue distribution of endogenous exosomes released from cardiomyocytes were demonstrated by luciferase assay and non-invasive bioluminescent live imaging. This endogenous exosome tracking mouse provides a useful tool for a range of research applications.With the development of data mining, machine learning offers opportunities to improve discrimination by analyzing complex interactions among massive variables. To test the ability of machine learning algorithms for predicting risk of type 2 diabetes mellitus (T2DM) in a rural Chinese population, we focus on a total of 36,652 eligible participants from the Henan Rural Cohort Study. Risk assessment models for T2DM were developed using six machine learning algorithms, including logistic regression (LR), classification and regression tree (CART), artificial neural networks (ANN), support vector machine (SVM), random forest (RF) and gradient boosting machine (GBM). The model performance was measured in an area under the receiver operating characteristic curve, sensitivity, specificity, positive predictive value, negative predictive value and area under precision recall curve. The importance of variables was identified based on each classifier and the shapley additive explanations approach. Using all available variables, all models for predicting risk of T2DM demonstrated strong predictive performance, with AUCs ranging between 0.811 and 0.872 using laboratory data and from 0.767 to 0.817 without laboratory data. Among them, the GBM model performed best (AUC 0.872 with laboratory data and 0.817 without laboratory data). Performance of models plateaued when introduced 30 variables to each model except CART model. Among the top-10 variables across all methods were sweet flavor, urine glucose, age, heart rate, creatinine, waist circumference, uric acid, pulse pressure, insulin, and hypertension. New important risk factors (urinary indicators, sweet flavor) were not found in previous risk prediction methods, but determined by machine learning in our study. Through the results, machine learning methods showed competence in predicting risk of T2DM, leading to greater insights on disease risk factors with no priori assumption of causality.In a myriad of engineering situations, we often hope to establish a model which can acquire load conditions around structures through flow features detection. A data-driven method is developed to predict the pressure on a cylinder from velocity distributions in its wake flow. The proposed deep learning neural network is constituted with convolutional layers and fully-connected layers The convolutional layers can process the velocity information by features extraction, which are gathered by the fully-connected layers to obtain the pressure coefficients. By comparing the output data of the typical network with Computational Fluid Dynamics (CFD) results as reference values, it suggests that the present convolutional neural network (CNN) is able to predict the pressure coefficient in the vicinity of the trained Reynolds numbers with various inlet flow profiles and achieves a high overall precision. Moreover, a transfer learning approach is adopted to preserve the feature detection ability by keeping the parameters in the convolutional layers unchanged while shifting parameters in the fully-connected layers. Further results show that this transfer learning network has nearly the same precision while significantly lower cost. The active prospects of convolutional neural network in fluid mechanics have also been demonstrated, which can inspire more kinds of loads prediction in the future.A case-controlled study was performed to evaluate taste and smell impairment, nausea or vomiting (NV) response to taste and smell and toleration to food texture, item and cooking method in hyperemesis gravidarum patients (HG) compared to gestation-matched controls from a university hospital and primary care clinic in Malaysia. Taste strips (4 base tastes), sniff sticks (16 selected smells) and a food-related questionnaire were used. 124 participants were recruited. Taste impairment was found in 13%(8/62) vs. selleck chemicals 0%(0/62) P = 0.003 and the median for correct smell identification was 5[4-6] vs. 9[7-9] P  less then  0.001 in HG vs. controls. In HG, bitter was most likely (32%) and sweet taste least likely (5%) to provoke NV. In both arms, fish smell was most likely to provoke NV, 77% vs. 32% P  less then  0.001 and peppermint smell least likely 10% vs. 0% P = 0.012; NV response was significantly more likely for HG arm in 10/16 smells. In HG, worst and best NV responses to food-texture were pasty 69% and crunchy 26%; food-item, plain rice 71% and apple 16% and cooking-style, deep-frying 71% and steaming 55%.