No abstract

In this paper, we introduce a domain-based random forest of decision trees to infer protein interactions. Our proposed method is capable of exploring all possible domain interactions and making predictions based on all the protein domains. Experimental results on Saccharomyces cerevisiae dataset demonstrate that our approach can predict protein-protein interactions with higher sensitivity (79.78%) and specificity (64.38%) compared with that of the maximum likelihood approach. Furthermore, our model can be used to infer interactions not only for single-domain pairs but also for multiple domain pairs.

Liver metastases develop in more than half of patients with colorectal cancer (CRC) and are associated with a poor prognosis. The factors influencing liver metastasis of CRC are poorly characterized, but this information is urgently needed. We have now discovered that small extracellular vesicles (sEVs, exosomes) derived from CRC can be specifically targeted to liver tissue and induce liver macrophage polarization toward an interleukin-6 (IL-6)-secreting pro-inflammatory phenotype. More importantly, we found that microRNA-21-5p (miR-21) was highly enriched in CRC-derived sEVs and was essential for creating a liver pro-inflammatory phenotype and liver metastasis of CRC. Silencing either miR-21 in CRC-sEVs or Toll-like receptor 7 (TLR7) in macrophages, to which miR-21 binds, abolished CRC-sEVs' induction of pro-inflammatory macrophages. Furthermore, miR-21 expression in plasma-derived sEVs was positively correlated with liver metastasis in CRC patients. Collectively, our data demonstrate a pivotal role of CRC-sEVs in promoting liver metastasis by inducing an inflammatory pre-metastatic niche through the miR-21-TLR7-IL6 axis. Thus, sEVs-miR-21 represents a potential prognostic marker and therapeutic target for CRC patients with liver metastasis.

No abstract

Auger recombination is the main non-radiative decay pathway for multi-carrier states of colloidal quantum dots, which affects performance of most of their optical and optoelectronic applications. Outstanding single-exciton properties of CdSe/CdS core/shell quantum dots enable us to simultaneously study the two basic types of Auger recombination channels—negative trion and positive trion channels. Though Auger rates of positive trion are regarded to be much faster than that of negative trion for II-VI quantum dots in literature, our experiments find the two rates can be inverted for certain core/shell geometries. This is confirmed by theoretical calculations as a result of geometry-dependent dielectric screening. By varying the core/shell geometry, both types of Auger rates can be independently tuned for ~ 1 order of magnitude. Experimental and theoretical findings shed new light on designing quantum dots with necessary Auger recombination characteristics for high-power light-emitting-diodes, lasers, single-molecular tracking, super-resolution microscope, and advanced quantum light sources.

The purpose of this investigation was to study the effect of the presence of red blood cells (RBCs) in the plasma layer near the arteriole wall on nitric oxide (NO) and oxygen (O2) transport. To this end, we extended a coupled NO and O2 diffusion-reaction model in the arteriole, developed by our group, to include the effect of the presence of RBCs in the plasma layer and the effect of convection. Two blood flow velocity profiles (plug and parabolic) were tested. The average hematocrit in the bloodstream was assumed to be constant in the central core and decreasing to zero in the boundary layer next to the endothelial surface layer. The effect of the presence or absence of RBCs near the endothelium was studied while varying the endothelial surface layer and boundary layer thickness. With RBCs present in the boundary layer, the model predicts that 1) NO decreases significantly in the endothelium and vascular wall; 2) there is a very small increase in endothelial and vascular wall Po2; 3) scavenging of NO by hemoglobin decreases with increasing thickness of the boundary layer; 4) the shape of the velocity profile influences both NO and Po2 gradients in the bloodstream; and 5) the presence of RBCs in the boundary layer near the endothelium has a much larger effect on NO than on O2 transport.

Large datasets often have unreliable labels-such as those obtained from Amazon's Mechanical Turk or social media platforms-and classifiers trained on mislabeled datasets often exhibit poor performance. We present a simple, effective technique for accounting for label noise when training deep neural networks. We augment a standard deep network with a softmax layer that models the label noise statistics. Then, we train the deep network and noise model jointly via endto-end stochastic gradient descent on the (perhaps mislabeled) dataset. The augmented model is overdetermined, so in order to encourage the learning of a non-trivial noise model, we apply dropout regularization to the weights of the noise model during training. Numerical experiments on noisy versions of the CIFAR-10 and MNIST datasets show that the proposed dropout technique outperforms state-of-the-art methods.The previous decade has witnessed swift advances in the performance of deep neural networks for supervised image classification and recognition. State-of-the-art performance requires large datasets, such as the 10,000,000 hand-labeled images comprising the ImageNet dataset [1], [2]. Large datasets suffer from noise, not only in the images themselves, but also in their associated labels. Researchers often resort to non-expert sources such as Amazon's Mechanical Turk or tags from social networking sites to label massive datasets, resulting in unreliable labels. Furthermore, the distinction between class labels is not always precise, and even experts may disagree on the correct label of an image. Regardless its source, the resulting noise can drastically degrade learning performance [3], [4].Learning with noisy labels has been studied previously, but not extensively. Techniques for training support vector machines, K-nearest neighbor classifiers, and logistic regression models with label noise are presented in [5], [6]. Further, [6] gives sample complexity bounds in the presence of label noise. Only a few papers consider deep learning with noisy labels. An early work is [7], which studied symmetric label noise in neural networks. Binary classification with label noise was studied in [8]. In [9], techniques for multi-class learning and general label noise models are presented. This approach adds an extra linear layer, intended to model the label noise, to the conventional convolutional neural network (CNN) architecture. In a similar vein, the work of [10] uses self-learning techniques to "bootstrap" the simultaneous learning of a deep network and a label noise model.In this paper, we present a simple, effective approach to learning deep neural networks from datasets corrupted by label flips. We augment an arbitrary deep architecture with a softmax layer that characterizes the pairwise label flip probabilities. We learn jointly the parameters of the deep network and the noise model simultaneously using standard stochastic gradient descent. To ensure that the network learns an accurate noise model-instead of fitting the deep network to the noisy labels...

Datasets and software are available at http://ittc.ku.edu/~xwchen/bindingsite/prediction.