Key challenges in proteomics and proteoinformatics

In this article, a survey on experimental and computational approaches related to proteomics is presented. Considered broadly, proteomics includes: techniques for identifying proteins in a sample, detecting posttranslational modifications (changes to proteins after translation), predicting the structure and function of proteins from sequence data, and integrating information about protein sequences from different databases. The paper focuses on the ways in which recent biological findings complicate the mapping from genes to RNA to protein. The authors argue that the challenges encountered in proteomics provide a valuable lesson on the complexity of life itself, as live organisms always contradict oversimplified models of biological information flow. In this overview, a snapshot of contemporary issues in proteomics is shown.     

Enabling proteomics discovery through visual analysis

This article presents the motivation for developing visual analysis tools for proteomic data and demonstrates their application to proteomics research with a visualization tool named Peptide Permutation and Protein Prediction, or PQuad, a functioning visual analytic tool for the study of systems biology, is in operation at the Pacific Northwest National Laboratory (PNNL). PQuad supports the exploration of proteins identified by proteomic techniques in the context of supplemental biological information. In particular, PQuad supports differential proteomics by simplifying the comparison of peptide sets from different experimental conditions as well as different proteins identification or confidence scoring techniques. Finally, PQuad supports data validation and quality control by providing a variety of resolutions for huge amounts of data to reveal errors undetected by other methods.     

Does protein structure influence trypsin miscleavage?

In this article, a new database that relates structural information from proteins in protein data bank to closely related protein sequences in humans was developed. Because the match criteria are extremely stringent, the structure of proteins in other species to infer characteristics of the human proteins was used. As a demonstration of the approach, this database has been applied to the problem of identifying likely trypsin miscleavage sites, a significant problem in proteomics. However, the approach is very general, and can be used to answer many kinds of structural questions (including questions related to posttranslational modifications). The study found that both the surface area and the secondary structure of cleavage sites have highly statistically significant effects on trypsin cleavage. The results of this analysis do not, however, suggest that surface area or secondary structure properties of particular peptides can be used to predict miscleavage sites, at least at a global level. This analysis of cleavage sites demonstrates the general power of homology-based techniques, in which the characteristics of a single protein that has a structure that has been solved can be used to infer properties of other proteins. We expect that our database of related proteins, structures, and sequences and our ability to query experimentally determined sets of peptides against this database will allow us to answer many other questions relation to global protein expression and modification.     

Enabling proteomics discovery through visual analysis. The peptide permutation and protein prediction tool.

Proteins play a key role in cellular processes, making proteomics central to understanding systems biology. MS techniques provide a means to observe entire proteomes at a global level. Yet, high-throughput MS proteomics techniques generate data faster than it can currently be analyzed. The success of proteomics depends on high-throughput experimental techniques coupled with sophisticated visual analysis and data-mining methods. Visual analysis has been applied successfully in a number of fields plagued with huge, complex data sets and will likely be an important tool in proteomics discovery. PQuad, a novel visualization of MS proteomics data, provides powerful analysis capabilities that support a number of proteomic data applications. In particular, PQuad supports differential proteomics by simplifying the comparison of peptide sets from different experimental conditions as well as different protein identification or confidence scoring techniques. Finally, PQuad supports data validation and quality control by providing a variety of resolutions for huge amounts of data to reveal errors undetected by other methods.     

Bioinformatics

Computational methods are becoming an increasingly important aspect of the evaluation and analysis of experimental data in molecular biology. The use of computational methods towards solving problems in biology is known as bioinformatics. The field of bioinformatics is constantly redefining itself as methods for collecting biological data are developed and refined. While the future directions of the field are impossible to predict, one conclusion seems to be evident: computational techniques have changed the way in which biologists collect and analyze experimental data. Computation will continue to be a prominent component of biochemistry and molecular biology research for the foreseeable future. While early studies developed the techniques necessary to sequence entire genomes, scientists are now investigating the interacting mechanisms that control the expression of genes. Ambitious new efforts are underway to identify the complex biological pathways of interaction between genes, the proteins for which they code, and the various metabolic intermediates acted upon by these proteins. Advances in understanding these sorts of large scale biological problems bear enormous promise for improving the human condition.     

基于WiFi指纹的层级学习室内定位模型

随着物联网和信息技术的飞速发展,基于移动位置的服务近年来日益受到关注,同时也促进了室内定位技术的发展.基于WiFi指纹的室内定位技术以其部署广泛、成本低廉等优点受到了学术界的广泛研究.针对移动设备在室内环境中的定位问题,提出了一种层级学习室内定位系统(hierarchical deep learning indoor ...     

Naïve Bayes Algorithm for Large Scale Text Classification

This paper proposed an improved Naïve Bayes Classifier for sentimental analysis from a large-scale dataset such as in YouTube. YouTube contains large unstructured and unorganized comments and reactions, which carry important in-formation. Organizing large amounts of data and extracting useful information is a challenging task. The extracted information can be considered as new knowledge and can be used for decision-making. We extract comments from YouTube on videos and categorized them in domain-specific, and then apply the Naïve Bayes classifier with improved techniques. Our method provided a decent 80% accuracy in classifying those comments. This experiment shows that the proposed method provides excellent adaptability for large-scale text classification.     

计算机网络中的故障定位技术研究

故障定位就是从一系列观测的故障现象中演绎出确切的故障源,也是网络故障管理的核心.一直以来有很多从计算机科学的不同领域衍生出来的方法解决故障定位,包括人工智能、图论、神经网络、信息论和自动机理论.本文论述了故障定位技术在计算机网络中的重要性,详细介绍了专家系统和图论方法在故障定位技术中的应用,特别论述了信度网故障定位方法.最后分析了目前故障定位尚待解决的问题.     

Genomics and Proteomics Engineering in Medicine and Biology (M. Akay, Ed.; 2007) [Book Reviews]

This book presents a well-rounded, interdisciplinary discussion of genomics and proteomics engineering. Compiling contributions by established experts, the book highlights up-to-date applications of biomedical informatics, as well as advancements in the genomics and proteomics areas. Topics discussed include: qualitative data models; interpreting microarray data; gene regulation bioinformatics; methods to analyze microarray; cancer behavior and radiation therapy; error-control codes and the genome complex life science multidatabase queries; computational protein analysis; and tumor and tumor suppressor proteins interactions.     

基于混合模型的超短期风速区间预测

准确的风速预测能够促进大规模的风电并网,保证电力系统的安全稳定运行。针对传统点预测方法难以表征预测结果概率可信度问题,提出一种基于模糊信息粒化、改进长短期记忆网络与差分自回归移动平均模型的混合区间预测模型。首先,采用自适应噪声的完全集合经验模态分解模型对原始风速数据进行分解,并依据模糊熵重构得到新序列。在此基础上,对每个序列依次进行模糊信息粒化,获得最大值、最小值及平均值。最后,利用改进长短期记忆网络模型预测高频序列,差分自回归移动平均模型预测低频序列与余项,并将所得上下界求和得到最终风速区间。算例分析表明,所提模型得出的风速预测区间能够准确覆盖实测风速,为电力系统调度提供更多有价值的决策信息。     

基于消息队列技术的建筑能耗监测系统应用研究

针对建筑能耗监测平台多系统集成带来的各种数据采集设备工作环境复杂和数据传输不同步的问题,提出了采用基于消息队列的技术提高数据采集系统的数据传输稳定性和可靠性的改进方案。构建了基于消息队列技术的建筑能耗数据采集系统,并以采集数据信息的消息发送和接收为例,给出了消息队列技术在建筑能耗数据采集系统中的核心源码,并对其在防止数据丢失和信息安全方面的作用作了总结。     

Engineering in genomics

This work addresses two principles that will be integral to the post-genomic or proteomic era (i.e., after sequencing). The first is that any analysis of data from or related to the Human Genome Project will need to be designed with high-throughput in mind. Just the sequence information will encompass some 3 billion nucleotides, and that does not include information about introns, exons, promoters, and many other features of interest. The volume of information that must be synthesized is even larger than the genome itself, and it is diverse in nature. It includes sequence, structural, functional, and localization information for each gene, and each of those constituents has its own levels of organization as well (e.g., functional information for a protein can be obtained at the molecular, cellular, and organismal levels). Computational analysis must be able to handle all these data in a reasonable amount of time. The second principle, which has been alluded to here, is that analysis techniques must incorporate data from a variety of sources. Archiving and indexing of sequence data, for example, must include sequences from multiple organisms and from diseased and healthy states to be maximally useful. The other levels of information, including structure, function, and localization will need to be similarly organized     

基于GIS的无线传感网架构及在轨道监测的应用

研究了基于GIS的无线传感网架构方法,建立了适用于大规模广域无线传感网的软件体系,开发了GIS服务、数据库服务、可视化与分析服务,形成了数据采集、汇聚、存储、与地理数据的融合、展示的完整流程,实现了无线采集的海量数据与GIS平台的动态融合,用户可以通过计算机、安卓智能手机等各种网络化终端设备由Internet访问传感系统,获取实时采集数据。最后对基于GIS的无线传感网系统进行了性能综合测试,结果表明系统能满足铁路轨道广域快速监测等重大工程需求。     

基于聚类技术的光伏并网影响量化研究

大型光伏系统并网对电网性能的影响分析是安装大型光伏系统前的一项重要工作。可以采用基于时间模拟的方法进行影响量化分析,其优点是能够提供有关光伏系统发电量波动的信息,缺点是需要进行大量的分析和模拟,尤其是在采用以1 h为采样间隔的长历史数据时,不适用于影响研究,因此提出了一种有效利用数据的新方法。该方法利用聚类技术将具有相似特征的光伏功率输出段进行分组,每组选择1个有代表性的输出段,用以分析和模拟。该代表性的输出段可提供关于组内其他输出段的预期性能信息。最后,通过大型光伏系统的选址定容,验证了该方法的有效性和实用性。     

考虑分布式光伏和电动汽车接入的配电网空间负荷预测方法

针对大规模分布式光伏和电动汽车接入配电网对空间负荷预测影响的问题,提出一种考虑远景年屋顶分布式光伏饱和安装、大规模电动汽车参与V2G的城市配电网空间负荷预测方法。区分不同小区,依据相应的容积率和可利用率系数计算屋顶光伏饱和安装面积,结合历史辐射值数据计算光伏出力。基于改进型停车生成率模型预测停车需求,结合日行驶里程、停车特性和充放电策略,建立电动汽车V2G负荷预测模型,利用蒙特卡罗仿真得出V2G负荷时空分布情况。采用改进型负荷密度指标法,实现对考虑时序的配电网传统日负荷的预测。以某规划区为例,预测结果表明:屋顶分布式光伏和电动汽车V2G对配电网空间负荷预测结果影响较大,且对不同小区负荷影响的程度不同。     

Data mining in protein interactomics

In this article, protein interactomics, an emerging field that studies the total collection of proteins and intracellular protein interactions in an organism, i.e., the study of protein interactomes is introduced. Protein interactomics is concerned with all the expressed proteins in a given tissue or cell type and how proteins physically interact with, or bind to, one another in the protein interaction network. Protein interactomes can provide information about protein functional links and protein functional context not apparent from either protein sequence analysis or protein expression analysis. By studying protein interactomics, biologists can compile biological pathway models to understand functional roles of previously uncharacterized proteins and biological processes in different developmental and environmental conditions. The paper discussed new biological discovery opportunities by presenting six specific data mining challenges in protein interactomics - data generation, data representation, data cleansing, data integration, data analysis/visualization, and knowledge curation.     

超大容量金属氢化物-镍电池的储能机遇

当前金属氢化物-镍(MH/Ni)电池在电力储能领域具备一定的优势,如进行差异化定位,可实现大型电力储能领域的拓展应用。MH/Ni电池具有安全性好、低温性能好及循环寿命长的优势,适用于电力储能领域,特别是高寒地区。MH/Ni电池大型化结构技术取得突破,开发超大容量MH/Ni电池,拓展储能领域的市场已成为现实。     

基于语义关联特征的大型信息管理系统数据挖掘技术

为了提高大型信息管理系统的数据检索和挖掘能力,提出了一种基于语义关联特征提取的大型信息管理系统数据挖掘技术。构建云存储模型进行大型信息管理系统中大数据分布式存储设计,结合大数据信息流的特征重组方法进行信息管理系统的优化结构重组,在重组的信息管理系统拓扑结构中提取信息管理分布数据的语义关联维特征量,以语义关联特征量为训练样本集进行信息管理系统的集成调度和数据挖掘,采用模糊C均值算法进行大型信息管理系统中分布数据语义关联特征的自适应融合和聚类处理,采用特征压缩器进行大型信息管理系统的存储空间降维处理,提高目标数据挖掘能力和信息管理系统的自适应调度能力。仿真结果表明,采用该方法进行大型信息管理系统数据挖掘的准确性较好,语义关联聚类性较强,提高了对信息管理系统目标数据的检索和调度能力。     

Electricity market price spike analysis by a hybrid data model and feature selection technique

In a competitive electricity market, energy price forecasting is an important activity for both suppliers and consumers. For this reason, many techniques have been proposed to predict electricity market prices in the recent years. However, electricity price is a complex volatile signal owning many spikes. Most of electricity price forecast techniques focus on the normal price prediction, while price spike forecast is a different and more complex prediction process. Price spike forecasting has two main aspects: prediction of price spike occurrence and value. In this paper, a novel technique for price spike occurrence prediction is presented composed of a new hybrid data model, a novel feature selection technique and an efficient forecast engine. The hybrid data model includes both wavelet and time domain variables as well as calendar indicators, comprising a large candidate input set. The set is refined by the proposed feature selection technique evaluating both relevancy and redundancy of the candidate inputs. The forecast engine is a probabilistic neural network, which are fed by the selected candidate inputs of the feature selection technique and predict price spike occurrence. The efficiency of the whole proposed method for price spike occurrence forecasting is evaluated by means of real data from the Queensland and PJM electricity markets.     

Distributed sensing and cooperativeestimation/detection of ubiquitous powerinternet of things

The rapid development of Internet Plus Smart Energy requires further strengthening of three kinds of interconnections based on traditional power systems: physical interconnection, information interconnection, and commercial interconnection. Due to the integration of renewable energy, the reform of the electricity market, and the deployment of the Smart Grid, a large amount of data will be generated. Thus, it is necessary to establish a Ubiquitous Power Internet of Things (UPIoT) to realize connections among people and things, things and things, and people and people in power systems. This paper studies the concept and architecture of the UPIoT and indicates the deployment of the perception layer and network layer as the key to building UPIoT in the initial stage. As UPIoT tends to cover a wide area and produce massive and distributed data, signal processing and data analytics theories and techniques are needed to handle the data and observe the state of the large-scale system. Further studies on distributed sensing and cooperative estimation theories and techniques of UPIoT are also required. Finally, the application prospects of UPIoT and the directions for future research are discussed.