Xiang Zhang

Bio: Xiang Zhang is an academic researcher from Baylor College of Medicine. The author has contributed to research in topics: Medicine & Computer science. The author has an hindex of 154, co-authored 1733 publications receiving 117576 citations. Previous affiliations of Xiang Zhang include University of California, Berkeley & University of Texas MD Anderson Cancer Center.

Flexible and robust co-regularized multi-domain graph clustering

Abstract: Multi-view graph clustering aims to enhance clustering performance by integrating heterogeneous information collected in different domains. Each domain provides a different view of the data instances. Leveraging cross-domain information has been demonstrated an effective way to achieve better clustering results. Despite the previous success, existing multi-view graph clustering methods usually assume that different views are available for the same set of instances. Thus instances in different domains can be treated as having strict one-to-one relationship. In many real-life applications, however, data instances in one domain may correspond to multiple instances in another domain. Moreover, relationships between instances in different domains may be associated with weights based on prior (partial) knowledge. In this paper, we propose a flexible and robust framework, CGC (Co-regularized Graph Clustering), based on non-negative matrix factorization (NMF), to tackle these challenges. CGC has several advantages over the existing methods. First, it supports many-to-many cross-domain instance relationship. Second, it incorporates weight on cross-domain relationship. Third, it allows partial cross-domain mapping so that graphs in different domains may have different sizes. Finally, it provides users with the extent to which the cross-domain instance relationship violates the in-domain clustering structure, and thus enables users to re-evaluate the consistency of the relationship. Extensive experimental results on UCI benchmark data sets, newsgroup data sets and biological interaction networks demonstrate the effectiveness of our approach.

Lung mesenchymal cells elicit lipid storage in neutrophils that fuel breast cancer lung metastasis

Abstract: Acquisition of a lipid-laden phenotype by immune cells has been defined in infectious diseases and atherosclerosis but remains largely uncharacterized in cancer. Here, in breast cancer models, we found that neutrophils are induced to accumulate neutral lipids upon interaction with resident mesenchymal cells in the premetastatic lung. Lung mesenchymal cells elicit this process through repressing the adipose triglyceride lipase (ATGL) activity in neutrophils in prostaglandin E2-dependent and -independent manners. In vivo, neutrophil-specific deletion of genes encoding ATGL or ATGL inhibitory factors altered neutrophil lipid profiles and breast tumor lung metastasis in mice. Mechanistically, lipids stored in lung neutrophils are transported to metastatic tumor cells through a macropinocytosis-lysosome pathway, endowing tumor cells with augmented survival and proliferative capacities. Pharmacological inhibition of macropinocytosis significantly reduced metastatic colonization by breast tumor cells in vivo. Collectively, our work reveals that neutrophils serve as an energy reservoir to fuel breast cancer lung metastasis.

Feedback-driven self-assembly of symmetry-breaking optical metamaterials in solution

Abstract: Thermodynamically driven self-assembly offers a direct route to organize individual nanoscopic components into threedimensional structures over a large scale 1–3 . The most thermodynamically favourable configurations, however, may not be ideal for some applications. In plasmonics, for instance, nanophotonic constructs with non-trivial broken symmetries can display optical properties of interest, such as Fano resonance, but are usually not thermodynamically favoured 4 . Here, we present a self-assembly route with a feedback mechanism for the bottom-up synthesis of a new class of symmetry-breaking optical metamaterials. We self-assemble plasmonic nanorod dimers with a longitudinal offset that determines the degree of symmetry breaking and its electromagnetic response. The clear difference in plasmonic resonance profiles of nanorod dimers in different configurations enables high spectra selectivity. On the basis of this plasmonic signature, our self-assembly route with feedback mechanism promotes the assembly of desired metamaterial structures through selective excitation and photothermal disassembly of unwanted assemblies in solution. In this fashion, our method can selectively reconfigure and homogenize the properties of the dimer, leading to highly monodispersed aqueous metamaterials with tailored symmetries and electromagnetic responses. In the last decade, control over structural symmetries has led to novel materials properties and the prediction of potentially exciting applications. Optical metamaterials 5 , unlike conventional materials, can be designed to have negative refraction, rainbow trapping and nonlinear metal optics. However, particularly rich and exciting physics comes into play when symmetry-breaking mechanisms are introduced that have the potential to enable strong anisotropic plasmon hybridization 6 , leading to unusual phenomena including plasmon-induced transparency 7 , anti-Hermitian plasmonic antennas 8 and optical magnetism, and consequently negative-index metamaterials 9 . These intriguing properties can be used in a number of

Scattering of surface plasmon polaritons at abrupt surface interfaces: Implications for nanoscale cavities

Abstract: We have developed a rigorous mode matching approach for the exact semi-analytical analysis of surface plasmon propagation across non-uniform semi-infinite dielectric-metal interfaces. We address two key deficiencies of related approaches in the literature: firstly, we resolve issues of accuracy and convergence and secondly, while we focus on the analysis of two-dimensional problems, we present a framework for three-dimensional problems for the first time. Analytical derivations of coupling coefficients between guided and radiation modes allow an efficient scattering matrix formulation to describe general structures with multiple discontinuities. Studies of the reflection, transmission and radiation of surface plasmons incident on both dielectric and metallic surface discontinuities show a correspondence with an effective Fresnel description. We also model a surface plasmon Distributed Bragg Reflector (DBR) capable of reflecting between 80% and 90% of incident surface plasmon power. Radiation mode scattering ultimately limits the DBR’s reflection performance rather than the intrinsic absorption of the metal. Thus alternative plasmonic geometries that suppress radiation modes, such as gap and channel structures, could be superior for the design of strongly reflective DBRs for integration in high Q-factor nano-scale cavities. We anticipate that this method will be an invaluable tool for the efficient and intuitive design of plasmonic devices based on structural non-uniformities.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers—8× deeper than VGG nets [40] but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions1, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Very Deep Convolutional Networks for Large-Scale Image Recognition

Abstract: In this work we investigate the effect of the convolutional network depth on its accuracy in the large-scale image recognition setting. Our main contribution is a thorough evaluation of networks of increasing depth using an architecture with very small (3x3) convolution filters, which shows that a significant improvement on the prior-art configurations can be achieved by pushing the depth to 16-19 weight layers. These findings were the basis of our ImageNet Challenge 2014 submission, where our team secured the first and the second places in the localisation and classification tracks respectively. We also show that our representations generalise well to other datasets, where they achieve state-of-the-art results. We have made our two best-performing ConvNet models publicly available to facilitate further research on the use of deep visual representations in computer vision.

Deep Residual Learning for Image Recognition

Abstract: Deeper neural networks are more difficult to train. We present a residual learning framework to ease the training of networks that are substantially deeper than those used previously. We explicitly reformulate the layers as learning residual functions with reference to the layer inputs, instead of learning unreferenced functions. We provide comprehensive empirical evidence showing that these residual networks are easier to optimize, and can gain accuracy from considerably increased depth. On the ImageNet dataset we evaluate residual nets with a depth of up to 152 layers---8x deeper than VGG nets but still having lower complexity. An ensemble of these residual nets achieves 3.57% error on the ImageNet test set. This result won the 1st place on the ILSVRC 2015 classification task. We also present analysis on CIFAR-10 with 100 and 1000 layers. The depth of representations is of central importance for many visual recognition tasks. Solely due to our extremely deep representations, we obtain a 28% relative improvement on the COCO object detection dataset. Deep residual nets are foundations of our submissions to ILSVRC & COCO 2015 competitions, where we also won the 1st places on the tasks of ImageNet detection, ImageNet localization, COCO detection, and COCO segmentation.

Going deeper with convolutions

Abstract: We propose a deep convolutional neural network architecture codenamed Inception that achieves the new state of the art for classification and detection in the ImageNet Large-Scale Visual Recognition Challenge 2014 (ILSVRC14). The main hallmark of this architecture is the improved utilization of the computing resources inside the network. By a carefully crafted design, we increased the depth and width of the network while keeping the computational budget constant. To optimize quality, the architectural decisions were based on the Hebbian principle and the intuition of multi-scale processing. One particular incarnation used in our submission for ILSVRC14 is called GoogLeNet, a 22 layers deep network, the quality of which is assessed in the context of classification and detection.