A maximum likelihood methodology for clusterwise linear regression

This paper presents a conditional mixture, maximum likelihood methodology for performing clusterwise linear regression. This new methodology simultaneously estimates separate regression functions and membership inK clusters or groups. A review of related procedures is discussed with an associated critique. The conditional mixture, maximum likelihood methodology is introduced together with the E-M algorithm utilized for parameter estimation. A Monte Carlo analysis is performed via a fractional factorial design to examine the performance of the procedure. Next, a marketing application is presented concerning the evaluations of trade show performance by senior marketing executives. Finally, other potential applications and directions for future research are identified.     

Investigation of proportional link linkage clustering methods

Proportional link linkage (PLL) clustering methods are a parametric family of monotone invariant agglomerative hierarchical clustering methods. This family includes the single, minimedian, and complete linkage clustering methods as special cases; its members are used in psychological and ecological applications. Since the literature on clustering space distortion is oriented to quantitative input data, we adapt its basic concepts to input data with only ordinal significance and analyze the space distortion properties of PLL methods. To enable PLL methods to be used when the numbern of objects being clustered is large, we describe an efficient PLL algorithm that operates inO(n ² logn) time andO(n ²) space.This work was partially supported by the Natural Sciences and Engineering Research Council of Canada and by the Austrian Fonds zur Förderung der wissenschaftlichen Forschung.     

The teachers/practitioners corner comparative analysis of load forecasting techniques at a southern utility

The prime directive of any regulated electric utility is to provide adequate and reliable electricity supplies to the consuming public at reasonable cost. This requires the continual addition of new generating plants which is based on a long term forecast of energy and peak demand. This study documents the forecasting process used at a southern utility and compares the accuracy of their models to that produced using Holt's exponential smoothing and generalized adoptive filtering.     

A common model approach to macroeconomics: using panel data to reduce sampling error

Is there a common model inherent in macroeconomic data? Macroeconomic theory suggests that market economies of various nations should share many similar dynamic patterns; as a result, individual country empirical models, for a wide variety of countries, often include the same variables. Yet, empirical studies often find important roles for idiosyncratic shocks in the differing macroeconomic performance of countries. We use forecasting criteria to examine the macrodynamic behaviour of 15 OECD countries in terms of a small set of familiar, widely used core economic variables, omitting country‐specific shocks. We find this small set of variables and a simple VAR ‘common model’ strongly support the hypothesis that many industrialized nations have similar macroeconomic dynamics. Copyright © 2005 John Wiley & Sons, Ltd.     

Structure of the Sec13/31 COPII coat cage

Endomembranes of eukaryotic cells are dynamic structures that are in continuous communication through the activity of specialized cellular machineries, such as the coat protein complex II (COPII), which mediates cargo export from the endoplasmic reticulum (ER). COPII consists of the Sar1 GTPase, Sec23 and Sec24 (Sec23/24), where Sec23 is a Sar1-specific GTPase-activating protein and Sec24 functions in cargo selection, and Sec13 and Sec31 (Sec13/31), which has a structural role. Whereas recent results have shown that Sec23/24 and Sec13/31 can self-assemble to form COPII cage-like particles, we now show that Sec13/31 can self-assemble to form minimal cages in the absence of Sec23/24. We present a three-dimensional reconstruction of these Sec13/31 cages at 30 A resolution using cryo-electron microscopy and single particle analysis. These results reveal a novel cuboctahedron geometry with the potential to form a flexible lattice and to generate a diverse range of containers. Our data are consistent with a model for COPII coat complex assembly in which Sec23/24 has a non-structural role as a multivalent ligand localizing the self-assembly of Sec13/31 to form a cage lattice driving ER cargo export.     

Insights from the genome of the biotrophic fungal plant pathogen Ustilago maydis

Ustilago maydis is a ubiquitous pathogen of maize and a well-established model organism for the study of plant-microbe interactions. This basidiomycete fungus does not use aggressive virulence strategies to kill its host. U. maydis belongs to the group of biotrophic parasites (the smuts) that depend on living tissue for proliferation and development. Here we report the genome sequence for a member of this economically important group of biotrophic fungi. The 20.5-million-base U. maydis genome assembly contains 6,902 predicted protein-encoding genes and lacks pathogenicity signatures found in the genomes of aggressive pathogenic fungi, for example a battery of cell-wall-degrading enzymes. However, we detected unexpected genomic features responsible for the pathogenicity of this organism. Specifically, we found 12 clusters of genes encoding small secreted proteins with unknown function. A significant fraction of these genes exists in small gene families. Expression analysis showed that most of the genes contained in these clusters are regulated together and induced in infected tissue. Deletion of individual clusters altered the virulence of U. maydis in five cases, ranging from a complete lack of symptoms to hypervirulence. Despite years of research into the mechanism of pathogenicity in U. maydis, no 'true' virulence factors had been previously identified. Thus, the discovery of the secreted protein gene clusters and the functional demonstration of their decisive role in the infection process illuminate previously unknown mechanisms of pathogenicity operating in biotrophic fungi. Genomic analysis is, similarly, likely to open up new avenues for the discovery of virulence determinants in other pathogens.     

Structural basis of cell surface receptor recognition by botulinum neurotoxin B

Botulinum neurotoxins (BoNTs) are potent bacterial toxins that cause paralysis at femtomolar concentrations by blocking neurotransmitter release. A 'double receptor' model has been proposed in which BoNTs recognize nerve terminals via interactions with both gangliosides and protein receptors that mediate their entry. Of seven BoNTs (subtypes A-G), the putative receptors for BoNT/A, BoNT/B and BoNT/G have been identified, but the molecular details that govern recognition remain undefined. Here we report the crystal structure of full-length BoNT/B in complex with the synaptotagmin II (Syt-II) recognition domain at 2.6 A resolution. The structure of the complex reveals that Syt-II forms a short helix that binds to a hydrophobic groove within the binding domain of BoNT/B. In addition, mutagenesis of amino acid residues within this interface on Syt-II affects binding of BoNT/B. Structural and sequence analysis reveals that this hydrophobic groove is conserved in the BoNT/G and BoNT/B subtypes, but varies in other clostridial neurotoxins. Furthermore, molecular docking studies using the ganglioside G(T1b) indicate that its binding site is more extensive than previously proposed and might form contacts with both BoNT/B and synaptotagmin. The results provide structural insights into how BoNTs recognize protein receptors and reveal a promising target for blocking toxin-receptor recognition.