Using ecological networks to answer questions in global biogeography and ecology

Ecological networks have classically been studied at site and landscape scales, yet recent efforts have been made to collate these data into global repositories. This offers an opportunity to integrate and upscale knowledge about ecological interactions from local to global scales to gain enhanced insights from the mechanistic information provided by these data. By drawing on existing research investigating patterns in ecological interactions at continental to global scales, we show how data on ecological networks, collected at appropriate scales, can be used to generate an improved understanding of many aspects of ecology and biogeography—for example, species distribution modelling, restoration ecology and conservation. We argue that by understanding the patterns in the structure and function of ecological networks across scales, it is possible to enhance our understanding of the natural world.     

ProDCoNN: Protein design using a convolutional neural network

Designing protein sequences that fold to a given three-dimensional (3D) structure has long been a challenging problem in computational structural biology with significant theoretical and practical implications. In this study, we first formulated this problem as predicting the residue type given the 3D structural environment around the C _α atom of a residue, which is repeated for each residue of a protein. We designed a nine-layer 3D deep convolutional neural network (CNN) that takes as input a gridded box with the atomic coordinates and types around a residue. Several CNN layers were designed to capture structure information at different scales, such as bond lengths, bond angles, torsion angles, and secondary structures. Trained on a very large number of protein structures, the method, called ProDCoNN (protein design with CNN), achieved state-of-the-art performance when tested on large numbers of test proteins and benchmark datasets.     

改进GA-SVM在冠状动脉疾病诊断中的应用

改进的遗传算法（GA）自动优化支持向量机（SVM）参数,同步决策最优特征子集。新颖的分组多基因交叉技术保留了基因小组中的信息,而且允许后代继承更多的来自染色体的遗传信息。该算法促进可行解集中的高质量染色体信息交换,提高了解空间的搜索能力。实验结果说明：改进GA-SVM不仅可决策出与疾病相关的重要特征变量、优化SVM参数,而且可提升分类性能。与前馈BP神经网络及自适应模糊推理系统两种学习算法的比较表明,改进GA-SVM具有更好地表现。     

Comparative analysis of inner cavities and ligand migration in non-symbiotic AHb1 and AHb2

This study reports a comparative analysis of the topological properties of inner cavities and the intrinsic dynamics of non-symbiotic hemoglobins AHb1 and AHb2 from Arabidopsis thaliana. The two proteins belong to the 3/3 globin fold and have a sequence identity of about 60%. However, it is widely assumed that they have distinct physiological roles. In order to investigate the structure–function relationships in these proteins, we have examined the bis-histidyl and ligand-bound hexacoordinated states by atomistic simulations using in silico structural models. The results allow us to identify two main pathways to the distal cavity in the bis-histidyl hexacoordinated proteins. Nevertheless, a larger accessibility to small gaseous molecules is found in AHb2. This effect can be attributed to three factors: the mutation Leu35(AHb1) → Phe32(AHb2), the enhanced flexibility of helix B, and the more favorable energetic profile for ligand migration to the distal cavity. The net effect of these factors would be to facilitate the access of ligands, thus compensating the preference for the fully hexacoordination of AHb2, in contrast to the equilibrium between hexa- and pentacoordinated species in AHb1. On the other hand, binding of the exogenous ligand introduces distinct structural changes in the two proteins. A well-defined tunnel is formed in AHb1, which might be relevant to accomplish the proposed NO detoxification reaction. In contrast, no similar tunnel is found in AHb2, which can be ascribed to the reduced flexibility of helix E imposed by the larger number of salt bridges compared to AHb1. This feature would thus support the storage and transport functions proposed for AHb2. This article is part of a Special Issue entitled: Oxygen Binding and Sensing Proteins.     

GVC-Net: Global Vascular Context Network for Cerebrovascular Segmentation Using Sparse Labels

ObjectivesCerebrovascular disease is a serious threat to human health. Because of its high mortality and disability rate, early diagnosis and prevention are very important. The performance of existing cerebrovascular segmentation methods based on deep learning depends on the integrity of labels. However, manual labels are usually of low quality and poor connectivity at small blood vessels, which directly affects the cerebrovascular segmentation results.Material and methodIn this paper, we propose a new segmentation network to segment cerebral vessels from MRA images by using sparse labels. The long-distance dependence between vascular structures is captured by the global vascular context module, and the topology is constrained by the hybrid loss function to segment the cerebral vessels with good connectivity.ResultExperiments show that our method performed with a sensitivity, precision, dice similarity coefficient, intersection over union and centerline dice similarity coefficient of 61.24%, 75.58%, 67.66%, 51.13% and 83.79% respectively.ConclusionThe obtained results reveal that the proposed cerebrovascular segmentation network has better segmentation performance for cerebrovascular segmentation under sparse labels, and can suppress the noise of background to a certain extent.     

The phosphatidylinositol 3-kinase inhibitor wortmannin alters the metabolism of the Alzheimer's amyloid precursor protein

One of the hallmarks of Alzheimer's disease is the accumulation of senile plaques in brain, extracellular lesions comprised mostly of aggregates of the amyloid beta-peptide (Abeta). Abeta is proteolytically derived from the Alzheimer's amyloid precursor protein (APP). The generation of Abeta and nonamyloidogenic derivatives of APP involves utilization of alternative processing pathways and multiple subcellular compartments. To improve our understanding of the regulation of APP processing, we investigated the effects of wortmannin, a phosphatidylinositol 3-kinase (PI3-kinase) inhibitor, on APP processing. PI3-kinases form a multifaceted family of enzymes that represent converging points for multiple signal transduction pathways and also act as key regulators of vesicular trafficking. In N2a neuroblastoma cells expressing either wild-type APP or the "Swedish" familial Alzheimer's disease-associated mutant variant of APP, wortmannin treatment resulted in decreased release of both Abeta and soluble APPalpha. In parallel, full-length APP and both processed derivatives accumulated inside the cells. These effects were not present at nanomolar concentrations of wortmannin, but only at micromolar concentrations, implying the possible involvement of a recently described trans-Golgi network (TGN)-associated PI3-kinase that is resistant to nanomolar concentrations of the inhibitor, but sensitive to micromolar concentrations. All effects were reversible when the drug was removed from the cell culture medium. Given the suspected site of action of this novel PI3-kinase activity at the TGN, it is tempting to speculate that the unexpected increase in the levels of both intracellular soluble APPalpha and intracellular Abeta might be due to wortmannin-induced covesiculation of APP together with its respective secretase enzymes within the TGN, leading to the execution of alpha-, beta-, and gamma-secretase reactions.     

植物中参与活性氧调控的基因网络

植物体内活性氧（reactive oxygen species,ROS）是氧化还原反应的必然副产物,具极高的活性和毒性,从而对细胞产生毒害。同时,活性氧作为信号分子对很多生理过程诸如植物生长发育、细胞程序化死亡及生物和非生物胁迫应答起调控作用。植物中ROS双重作用的协调机制目前尚不明确,确定的是细胞中ROS维持于稳定水平需要精细的调节。拟南芥中至少包括152个基因组成的网络参与ROS的调控,该网络具高度的灵活性和互补性。本文综述了ROS网络中鉴定的一些关键基因及细胞学定位和协同作用,ROS信号转导,尤其是叶绿体中ROS信号的调控。     

All that glisters is not galled

Galled trees, evolutionary networks with isolated reticulation cycles, have appeared under several slightly different definitions in the literature. In this paper, we establish the actual relationships between the main four such alternative definitions: namely, the original galled trees, level-1 networks, nested networks with nesting depth 1, and evolutionary networks with arc-disjoint reticulation cycles.     

Fast and robust image segmentation by small-world neural oscillator networks

Inspired by the temporal correlation theory of brain functions, researchers have presented a number of neural oscillator networks to implement visual scene segmentation problems. Recently, it is shown that many biological neural networks are typical small-world networks. In this paper, we propose and investigate two small-world models derived from the well-known LEGION (locally excitatory and globally inhibitory oscillator network) model. To form a small-world network, we add a proper proportion of unidirectional shortcuts (random long-range connections) to the original LEGION model. With local connections and shortcuts, the neural oscillators can not only communicate with neighbors but also exchange phase information with remote partners. Model 1 introduces excitatory shortcuts to enhance the synchronization within an oscillator group representing the same object. Model 2 goes further to replace the global inhibitor with a sparse set of inhibitory shortcuts. Simulation results indicate that the proposed small-world models could achieve synchronization faster than the original LEGION model and are more likely to bind disconnected image regions belonging together. In addition, we argue that these two models are more biologically plausible.