Learning-aided predictor integration for system performance prediction

The integration of multiple predictors promises higher prediction accuracy than the accuracy that can be obtained with a single predictor. The challenge is how to select the best predictor at any given moment. Traditionally, multiple predictors are run in parallel and the one that generates the best result is selected for prediction. In this paper, we propose a novel approach for predictor integration based on the learning of historical predictions. Compared with the traditional approach, it does not require running all the predictors simultaneously. Instead, it uses classification algorithms such as k-Nearest Neighbor (k-NN) and Bayesian classification and dimension reduction technique such as Principal Component Analysis (PCA) to forecast the best predictor for the workload under study based on the learning of historical predictions. Then only the forecasted best predictor is run for prediction. Our experimental results show that it achieved 20.18% higher best predictor forecasting accuracy than the cumulative MSE based predictor selection approach used in the popular Network Weather Service system. In addition, it outperformed the observed most accurate single predictor in the pool for 44.23% of the performance traces.

Q-NET – a new scholarly network on quantitative wood anatomy

Quantitative wood anatomy (QWA) is a dynamic research approach of increasing interest that can provide answers to a wide range of research questions across different disciplines. However, the lack of common protocols and knowledge gaps hinder the realisation of the full potential of QWA. Therefore, we established the new community-based network Q-NET to provide an open interdisciplinary platform for exchange and research around QWA. Q-NET (https://qwa-net.com) combines an online knowledge and exchange base with virtual workshops. The first two workshops each attracted more than 125 participants from around the world, demonstrating the community's interest in QWA and this virtual way of networking and collaborating. Indeed, virtual networks such as Q-NET could increase the inclusiveness, efficiency and sustainability of scientific collaboration while providing additional training and teaching opportunities for early career scientists, both of which complement in-person conferences and workshops.     

Design of high performance nanozymes: a single-atom strategy

<正>Nanozymes, nanomaterials with enzyme-like characteristics,are emerging as novel artificial enzymes (Gao et al., 2007;Manea et al., 2004; Yan, 2018). They are superior to natural enzymes in many ways, such as higher stability, lower cost in preparation, and better robustness toward harsh environments (Wei and Wang, 2013). Various nanomaterials (e.g.,     

Gene regulation: hacking the network on a sugar high

In a recent issue of Molecular Cell, Lee et al. (2008) demonstrate that the forkhead-associated (FHA) domain of Dun1, a Chk2-related kinase involved in DNA damage signaling, selectively binds a diphosphorylated motif found in its paralog and upstream regulator Rad53.     

CCN proteins: A centralized communication network

The CCN family of proteins includes six members presently known as CCN1, CCN2, CCN3, CCN4, CCN5 and CCN6. These proteins were originally designated CYR61, CTGF, NOV, and WISP-1, WISP-2, WISP-3. Although these proteins share a significant amount of structural features and a partial identity with other large families of regulatory proteins, they exhibit different biological functions. A critical examination of the progress made over the past two decades, since the first CCN proteins were discovered brings me to the conclusion that most of our present knowledge regarding the functions of these proteins was predicted very early after their discovery. In an effort to point out some of the gaps that prevent us to reach a comprehensive view of the functional interactions between CCN proteins, it is necessary to reconsider carefully data that was already published and put aside, either because the scientific community was not ready to accept them, or because they were not fitting with the « consensus » when they were published. This review article points to avenues that were not attracting the attention that they deserved. However, it is quite obvious that the six members of this unique family of tetra-modular proteins must act in concert, either simultaneously or sequentially, on the same sites or at different times in the life of living organisms. A better understanding of the spatio-temporal regulation of CCN proteins expression requires considering the family as such, not as a set of single proteins related only by their name. As proposed in this review, there is enough convincing pieces of evidence, at the present time, in favor of these proteins playing a role in the coordination of multiple signaling pathways, and constituting a Centralized Communication Network. Deciphering the hierarchy of regulatory circuits involved in this complex system is an important challenge for the near future. In this article, I would like to briefly review the concept of a CCN family of proteins and critically examine the progress made over the past 10 years in the understanding of their biological functions and involvement in both normal and pathological processes.     

Extended performance using a high field magnet mass spectrometer

The development of a high field magnet for high mass electron impact, chemical ionization, field desorption and fast atom bombardment mass spectrometric studies is described. Its utility is illustrated with examples from structural studies of vitamin B12 biosynthetic intermediates, oligosaccharides, glycopeptides and the bleomycin antibiotics. The technique has also greatly assisted sequence studies of protein derived peptides.     

Application of immobilized enzyme reactor in on-line high performance liquid chromatography: a review

This review summarizes all the research efforts in the last decade (1994-2003) that have been spent to the various application of immobilized enzyme reactor (IMER) in on-line high performance liquid chromatography (HPLC). All immobilization procedures including supports, kind of assembly into chromatographic system and methods are described. The effect of immobilization on enzymatic properties and stability of biocatalysts is considered. A brief survey of the main applications of IMER both as pre-column, post-column or column in the chemical, pharmaceutical, clinical and commodities fields is also reported.     

Quorum sensing and fungal–bacterial interactions in Candida albicans: a communicative network regulating microbial coexistence and virulence

Microorganisms have evolved a complex signature of communication termed quorum sensing (QS), which is based on the exchange and sensing of low-molecular-weight signal compounds. The ability to communicate within the microbial population gives the advantage to coordinate a groups behaviour leading to a higher fitness in the environment. The polymorphic fungus Candida albicans is an opportunistic human pathogen able to regulate virulence traits through the production of at least two QS signal molecules: farnesol and tyrosol. The ability to adopt multiple morphotypes and form biofilms on infected surfaces are the most important pathogenic characteristics regulated by QS and are of clinical relevance. In fact, traditional antimicrobial approaches are often ineffective towards these characteristics. Moreover, the intimate association between C. albicans and other pathogens, such as Pseudomonas aeruginosa , increases the complexity of the infection system. This review outlines the current knowledge on fungal QS and fungal–bacterial interactions emphasizing on C. albicans . Further investigations need to concentrate on the molecular mechanisms and the genetic regulation of these phenomena in order to identify putative novel therapeutic options.     

The malignant social network

Tumors contain a vastly complicated cellular network that relies on local communication to execute malignant programs. The molecular cues that are involved in cell-cell adhesion orchestrate large-scale tumor behaviors such as proliferation and invasion. We have recently begun to appreciate that many tumors contain a high degree of cellular heterogeneity and are organized in a cellular hierarchy, with a cancer stem cell (CSC) population identified at the apex in multiple cancer types. CSCs reside in unique microenvironments or niches that are responsible for directing their behavior through cellular interactions between CSCs and stromal cells, generating a malignant social network. Identifying cell-cell adhesion mechanisms in this network has implications for the basic understanding of tumorigenesis and the development of more effective therapies. In this review, we will discuss our current understanding of cell-cell adhesion mechanisms used by CSCs and how these local interactions have global consequences for tumor biology.     

Design of a set of probes with high potential for influenza virus epidemiological surveillance

An Influenza Probe Set (IPS) consisting in 1,249 9-mer probes for genomic fingerprinting of closely and distantly related Influenza Virus strains was designed and tested in silico. The IPS was derived from alignments of Influenza genomes. The RNA segments of 5,133 influenza strains having diverse degree of relatedness were concatenated and aligned. After alignment, 9-mer sites having high Shannon entropy were searched. Additional criteria such as: G+C content between 35 to 65%, absence of dimer or trimer consecutive repeats, a minimum of 2 differences between 9mers and selecting only sequences with Tm values between 34.5 and 36.5oC were applied for selecting probes with high sequential entropy. Virtual Hybridization was used to predict Genomic Fingerprints to assess the capability of the IPS to discriminate between influenza and related strains. Distance scores between pairs of Influenza Genomic Fingerprints were calculated, and used for estimating Taxonomic Trees. Visual examination of both Genomic Fingerprints and Taxonomic Trees suggest that the IPS is able to discriminate between distant and closely related Influenza strains. It is proposed that the IPS can be used to investigate, by virtual or experimental hybridization, any new, and potentially virulent, strain.     

Design and performance of a 24-station high throughput microbioreactor

Two prototype 24-unit microbioreactors are presented and reviewed for their relative merits. The first used a standard 24-well plate as the template, while the second consisted of 24-discrete units. Both systems used non-invasive optical sensors to monitor pH and dissolved oxygen. The systems were used to cultivate Escherichia coli. Both designs had their merits and the results obtained are presented. In addition, dissolved oxygen control was demonstrated at the milliliter scale and 24 simultaneously monitored fermentations were successfully carried out. These results demonstrated high quality high throughput bioprocessing and provide important insights into operational parameters at small scale.     

The malignant social network: Cell-cell adhesion and communication in cancer stem cells

Tumors contain a vastly complicated cellular network that relies on local communication to execute malignant programs. The molecular cues that are involved in cell-cell adhesion orchestrate large-scale tumor behaviors such as proliferation and invasion. We have recently begun to appreciate that many tumors contain a high degree of cellular heterogeneity and are organized in a cellular hierarchy, with a cancer stem cell (CSC) population identified at the apex in multiple cancer types. CSCs reside in unique microenvironments or niches that are responsible for directing their behavior through cellular interactions between CSCs and stromal cells, generating a malignant social network. Identifying cell-cell adhesion mechanisms in this network has implications for the basic understanding of tumorigenesis and the development of more effective therapies. In this review, we will discuss our current understanding of cell-cell adhesion mechanisms used by CSCs and how these local interactions have global consequences for tumor biology.     

Spider silk aging: initial improvement in a high performance material followed by slow degradation

Spider silk possesses a unique combination of high tensile strength and elasticity resulting in extraordinarily tough fibers, compared with the best synthetic materials. However, the potential application of spider silk and biomimetic fibers depends upon retention of their high performance under a variety of conditions. Here, we report on changes in the mechanical properties of dragline and capture silk fibers from several spider species over periods up to 4 years of benign aging. We find an improvement in mechanical performance of silk fibers during the first year of aging. Fibers rapidly decrease in diameter, suggesting an increase in structural alignment and organization of molecules. One-year old silk also is stiffer and has higher stress at yield than fresh silk, whereas breaking force, elasticity, and toughness either improve or are unaffected by early aging. However, 4-year old silk shows signs of degradation as the breaking load, elasticity, and toughness are all lower than in fresh silk. Aging, however, does not reduce the tensile strength of silk. These data suggest initially rapid reorganization and tighter packaging of molecules within the fiber, followed by longer-term decomposition. We hypothesize that possibly the breakdown of amino acids via emission of ammonia gas, as is seen in long-term aging of museum silkworm fabrics, may contribute. Degradation of spider silk under benign conditions may be a concern for efforts to construct and utilize biomimetic silk analogs. However, our findings suggest an initial improvement in mechanical performance and that even old spider silk still retains impressive mechanical performance. J. Exp. Zool. 309A:494-504, 2008. (c) 2008 Wiley-Liss, Inc.     

The poly dA helix: a new structural motif for high performance DNA-based molecular switches

We report a pH-dependent conformational transition in short, defined homopolymeric deoxyadenosines (dA₁₅) from a single helical structure with stacked nucleobases at neutral pH to a double-helical, parallel-stranded duplex held together by AH⁺-H⁺A base pairs at acidic pH. Using native PAGE, 2D NMR, circular dichroism (CD) and fluorescence spectroscopy, we have characterized the two different pH dependent forms of dA₁₅. The pH-triggered transition between the two defined helical forms of dA₁₅ is characterized by CD and fluorescence. The kinetics of this conformational switch is found to occur on a millisecond time scale. This robust, highly reversible, pH-induced transition between the two well-defined structured states of dA₁₅ represents a new molecular building block for the construction of quick-response, pH-switchable architectures in structural DNA nanotechnology.     

Zeptosens' protein microarrays: a novel high performance microarray platform for low abundance protein analysis

Protein microarrays are considered an enabling technology, which will significantly expand the scope of current protein expression and protein interaction analysis. Current technologies, such as two-dimensional gel electrophoresis (2-DE) in combination with mass spectrometry, allowing the identification of biologically relevant proteins, have a high resolving power, but also considerable limitations. As was demonstrated by Gygi et al. (Proc. Nat. Acad. Sci. USA 2000,97, 9390-9395), most spots in 2-DE, observed from whole cell extracts, are from high abundance proteins, whereas low abundance proteins, such as signaling molecules or kinases, are only poorly represented. Protein microarrays are expected to significantly expedite the discovery of new markers and targets of pharmaceutical interest, and to have the potential for high-throughput applications. Key factors to reach this goal are: high read-out sensitivity for quantification also of low abundance proteins, functional analysis of proteins, short assay analysis times, ease of handling and the ability to integrate a variety of different targets and new assays. Zeptosens has developed a revolutionary new bioanalytical system based on the proprietary planar waveguide technology which allows us to perform multiplexed, quantitative biomolecular interaction analysis with highest sensitivity in a microarray format upon utilizing the specific advantages of the evanescent field fluorescence detection. The analytical system, comprising an ultrasensitive fluorescence reader and microarray chips with integrated microfluidics, enables the user to generate a multitude of high fidelity data in applications such as protein expression profiling or investigating protein-protein interactions. In this paper, the important factors for developing high performance protein microarray systems, especially for targeting low abundant messengers of relevant biological information, will be discussed and the performance of the system will be demonstrated in experimental examples.     

Timing of induced resistance in a clonal plant network

After local herbivory, plants can activate defense traits both at the damaged site and in undamaged plant parts such as in connected ramets of clonal plants. Since defense induction has costs, a mismatch in time and space between defense activation and herbivore feeding might result in negative consequences for plant fitness. A short time lag between attack and defense activation is important to ensure efficient protection of the plant. Additionally, the duration of induced defense production once the attack has stopped is also relevant in assessing the cost–benefit balance of inducible defenses, which will depend on the absence or presence of subsequent attacks. In this study we quantified the timing of induced responses in ramet networks of the stoloniferous herb Trifolium repens after local damage by Mamestra brassicae larvae. We studied the activation time of systemic defense induction in undamaged ramets and the decay time of the response after local attack. Undamaged ramets became defense‐induced 38–51 h after the initial attack. Defense induction was measured as a reduction in leaf palatability. Defense induction lasted at least 28 days, and there was strong genotypic variation in the duration of this response. Ramets formed after the initial attack were also defense‐induced, implying that induced defense can extend to new ramet generations, thereby contributing to protection of plant tissue that is both very vulnerable to herbivores and most valuable in terms of future plant growth and fitness.     

Members only: induced systemic resistance to herbivory in a clonal plant network

The stoloniferous herb Trifolium repens was used to study the expression of induced systemic resistance (ISR) to the generalist caterpillar Spodoptera exigua in interconnected ramets of clonal fragments. The ISR was assessed as caterpillar preference in dual choice tests between control and systemically induced plants. The ISR was detected in young ramets, after inducing older sibling ramets on the same stolon by a controlled herbivore attack. However, older ramets did not receive a defense induction signal from younger ramets unless the predominant phloem flow was reversed by means of basal shading. This provides evidence for the notion that in T. repens the clone-internal expression of ISR is coupled to phloem transport and follows source–sink gradients. The inducibility of the genotypes was not linked to their constitutive ability to produce cyanide, implying the absence of a trade-off between these two defense traits. To our knowledge, this is the first study that explores ISR to herbivory in the context of physiological integration in potentially extensive clonal plant networks.     

Trait‐based risk assessment for invasive species: high performance across diverse taxonomic groups,geographic ranges and machine learning/statistical tools

Aim Trait‐based risk assessment for invasive species is becoming an important tool for identifying non‐indigenous species that are likely to cause harm. Despite this, concerns remain that the invasion process is too complex for accurate predictions to be made. Our goal was to test risk assessment performance across a range of taxonomic and geographical scales, at different points in the invasion process, with a range of statistical and machine learning algorithms. Location Regional to global data sets. Methods We selected six data sets differing in size, geography and taxonomic scope. For each data set, we created seven risk assessment tools using a range of statistical and machine learning algorithms. Performance of tools was compared to determine the effects of data set size and scale, the algorithm used, and to determine overall performance of the trait‐based risk assessment approach. Results Risk assessment tools with good performance were generated for all data sets. Random forests (RF) and logistic regression (LR) consistently produced tools with high performance. Other algorithms had varied performance. Despite their greater power and flexibility, machine learning algorithms did not systematically outperform statistical algorithms. Geographic scope of the data set, and size of the data set, did not systematically affect risk assessment performance. Main conclusions Across six representative data sets, we were able to create risk assessment tools with high performance. Additional data sets could be generated for other taxonomic groups and regions, and these could support efforts to prevent the arrival of new invaders. Random forests and LR approaches performed well for all data sets and could be used as a standard approach to risk assessment development.