Cluster analysis of high-dimensional data: A case study

Normal mixture models are often used to cluster continuous data. However, conventional approaches for fitting these models will have problems in producing nonsingular estimates of the component-covariance matrices when the dimension of the observations is large relative to the number of observations. In this case, methods such as principal components analysis (PCA) and the mixture of factor analyzers model can be adopted to avoid these estimation problems. We examine these approaches applied to the Cabernet wine data set of Ashenfelter (1999), considering the clustering of both the wines and the judges, and comparing our results with another analysis. The mixture of factor analyzers model proves particularly effective in clustering the wines, accurately classifying many of the wines by location.

The ABC transporter gene in the sirodesmin biosynthetic gene cluster of Leptosphaeria maculans is not essential for sirodesmin production but facilitates self-protection

Epipolythiodioxopiperazine toxins are secreted by a range of fungi, including Leptosphaeria maculans, which produces sirodesmin, and Aspergillus fumigatus, which produces gliotoxin. The L. maculans biosynthetic gene cluster for sirodesmin includes an ABC transporter gene, sirA. Disruption of this gene led to increased secretion of sirodesmin into the medium and an altered ratio of sirodesmin to its immediate precursor. The transcription pattern of a peptide synthetase that catalyses an early step in sirodesmin biosynthesis was elevated in the sirA mutant by 47% over a 7-day period. This was consistent with the finding that the transporter mutant had elevated sirodesmin levels. Despite increased production of sirodesmin, the sit-A mutant was more sensitive to both sirodesmin and gliotoxin. The putative gliotoxin transporter gene, gliA, (a major facilitator superfamily transporter) from A.fumigatus complemented the tolerance of the L. maculans sirA mutant to gliotoxin, but not to sirodesmin. The results indicate that SirA contributes to self-protection against sirodesmin in L. maculans and suggest a transporter other than SirA is primarily responsible for efflux of endogenously produced sirodesmin. (C) 2004 Elsevier Inc. All rights reserved.

SARS coronavirus nucleocapsid immunodominant T-cell epitope cluster is common to both exogenous recombinant and endogenous DNA-encoded immunogens

Correspondence between the T-cell epitope responses of vaccine immunogens and those of pathogen antigens is critical to vaccine efficacy. In the present study, we analyzed the spectrum of immune responses of mice to three different forms of the SARS coronavirus nucleocapsid (N): (1) exogenous recombinant protein (N-GST) with Freund's adjuvant; (2) DNA encoding unmodified N as an endogenous cytoplasmic protein (pN); and (3) DNA encoding N as a LAMP-I chimera targeted to the lysosomal MHC II compartment (p-LAMP-N). Lysosomal trafficking of the LAMP/N chimera in transfected cells was documented by both confocal and immunoelectron microscopy. The responses of the immunized mice differed markedly. The strongest T-cell IFN-gamma and CTL responses were to the LAMP-N chimera followed by the pN immunogen. In contrast, N-GST elicited strong T cell IL-4 but minimal IFN-gamma responses and a much greater antibody response. Despite these differences, however, the immunodominant T-cell ELISpot responses to each of the three immunogens were elicited by the same N peptides, with the greatest responses being generated by a cluster of five overlapping peptides, N76-114, each of which contained nonameric H2(d) binding domains with high binding scores for both class I and, except for N76-93, class II alleles. These results demonstrate that processing and presentation of N, whether exogenously or endogenously derived, resulted in common immunodominant epitopes, supporting the usefulness of modified antigen delivery and trafficking forms and, in particular, LAMP chimeras as vaccine candidates. Nevertheless, the profiles of T-cell responses were distinctly different. The pronounced Th-2 and humoral response to N protein plus adjuvant are in contrast to the balanced IFN-gamma and IL-4 responses and strong memory CTL responses to the LAMP-N chimera. (C) 2005 Elsevier Inc. All rights reserved.

A pragmatic 2 × 2 × 2 factorial cluster randomized controlled trial of educational outreach visiting and case conferencing in palliative care—methodology of the Palliative Care Trial [ISRCTN 81117481]

The demand for palliative care is increasing, yet there are few data on the best models of care nor well-validated interventions that translate current evidence into clinical practice. Supporting multidisciplinary patient-centered palliative care while successfully conducting a large clinical trial is a challenge. The Palliative Care Trial (PCT) is a pragmatic 2 x 2 x 2 factorial cluster randomized controlled trial that tests the ability of educational outreach visiting and case conferencing to improve patient-based outcomes such as performance status and pain intensity. Four hundred sixty-one consenting patients and their general practitioners (GPs) were randomized to the following: (1) GP educational outreach visiting versus usual care, (2) Structured patient and caregiver educational outreach visiting versus usual care and (3) A coordinated palliative care model of case conferencing versus the standard model of palliative care in Adelaide, South Australia (3:1 randomization). Main outcome measures included patient functional status over time, pain intensity, and resource utilization. Participants were followed longitudinally until death or November 30, 2004. The interventions are aimed at translating current evidence into clinical practice and there was particular attention in the trial's design to addressing common pitfalls for clinical studies in palliative care. Given the need for evidence about optimal interventions and service delivery models that improve the care of people with life-limiting illness, the results of this rigorous, high quality clinical trial will inform practice. Initial results are expected in mid 2005. (c) 2005 Elsevier Inc. All rights reserved.

Cluster-state quantum computation

This article is a short introduction to and review of the cluster-state model of quantum computation, in which coherent quantum information processing is accomplished via a sequence of single-qubit measurements applied to a fixed quantum state known as a cluster state. We also discuss a few novel properties of the model, including a proof that the cluster state cannot occur as the exact ground state of any naturally occurring physical system, and a proof that measurements on any quantum state which is linearly prepared in one dimension can be efficiently simulated on a classical computer, and thus are not candidates for use as a substrate for quantum computation.

Community and cluster centre residential services for adults with intellectual disability: long-term results from an Australian-matched sample

Changes in residential accommodation models for adults with intellectual disability (ID) over the last 20 years in Australia, the United Kingdom and the United States have involved relocation from institutions primarily into dispersed homes in the community. But an evolving alternative service style is the cluster centre. This paper reports on the relocation of a matched group of 30 pairs of adults with moderate and severe IDs and challenging behaviour who were relocated from an institution into either dispersed housing in the community or cluster centres but under the same residential service philosophy. Adaptive and maladaptive behaviour, choice-making and objective life quality were assessed prior to leaving the institution and then after 12 and 24 months of living in the new residential model. Adaptive behaviour, choice-making and life quality increased for both groups and there was no change in level of maladaptive behaviour compared with levels exhibited in the institution. However, there were some significant differences between the community and cluster centre group as the community group increased some adaptive skills, choice-making and objective life quality to a greater extent than the cluster centre group. Both cluster centre and dispersed community living offer lifestyle and skill development advantages compared with opportunities available in large residential institutions. Dispersed community houses, however, offer increased opportunities for choice-making, acquisition of adaptive behaviours and improved life quality for long-term institutionalized adults with IDs.

C-terminal charged cluster of MscL, RKKEE, functions as a pH sensor

The RKKEE cluster of charged residues located within the cytoplasmic helix of the bacterial mechanosensitive channel, MscL, is essential for the channel function. The structure of MscL determined by x-ray crystallography and electron paramagnetic resonance spectroscopy has revealed discrepancies toward the C-terminus suggesting that the structure of the C-terminal helical bundle differs depending on the pH of the cytoplasm. In this study we examined the effect of pH as well as charge reversal and residue substitution within the RKKEE cluster on the mechanosensitivity of Escherichia coli MscL reconstituted into liposomes using the patch-clamp technique. Protonation of either positively or negatively charged residues within the cluster, achieved by changing the experimental pH or residue substitution within the RKKEE cluster, significantly increased the free energy of activation for the MscL channel due to an increase in activation pressure. Our data suggest that the orientation of the C-terminal helices relative to the aqueous medium is pH dependent, indicating that the RKKEE cluster functions as a proton sensor by adjusting the channel sensitivity to membrane tension in a pH-dependent fashion. A possible implication of our results for the physiology of bacterial cells is briefly discussed.

Discovery of ultracompact dwarf galaxies in the Virgo Cluster

We have discovered nine ultracompact dwarf galaxies (UCDs) in the Virgo Cluster, extending samples of these objects outside the Fornax Cluster. Using the Two Degree Field (2dF) multifiber spectrograph on the Anglo-Australian Telescope, the new Virgo members were found among 1500 color-selected, starlike targets with 16: 0 < b(j) < 20.2 in a 2 degrees diameter field centered on M87 (NGC 4486). The newly found UCDs are comparable to the UCDs in the Fornax Cluster, with sizes less than or similar to 100 pc, -12.9 < M-B < -10.7, and exhibiting red absorption-line spectra, indicative of an older stellar population. The properties of these objects remain consistent with the tidal threshing model for the origin of UCDs from the surviving nuclei of nucleated dwarf elliptical galaxies disrupted in the cluster core but can also be explained as objects that were formed by mergers of star clusters created in galaxy interactions. The discovery that UCDs exist in Virgo shows that this galaxy type is probably a ubiquitous phenomenon in clusters of galaxies; coupled with their possible origin by tidal threshing, the UCD population is a potential indicator and probe of the formation history of a given cluster. We also describe one additional bright UCD with M-B = -12.0 in the core of the Fornax Cluster. We find no further UCDs in our Fornax Cluster Spectroscopic Survey down to bj 19.5 in two additional 2dF fields extending as far as 3 degrees from the center of the cluster. All six Fornax bright UCDs identified with 2dF lie within 0.degrees 5 (projected distance of 170 kpc) of the central elliptical galaxy NGC 1399.

Noise thresholds for optical cluster-state quantum computation

In this paper we do a detailed numerical investigation of the fault-tolerant threshold for optical cluster-state quantum computation. Our noise model allows both photon loss and depolarizing noise, as a general proxy for all types of local noise other than photon loss noise. We obtain a threshold region of allowed pairs of values for the two types of noise. Roughly speaking, our results show that scalable optical quantum computing is possible in the combined presence of both noise types, provided that the loss probability is less than 3 X 10(-3) and the depolarization probability is less than 10(-4). Our fault-tolerant protocol involves a number of innovations, including a method for syndrome extraction known as telecorrection, whereby repeated syndrome measurements are guaranteed to agree. This paper is an extended version of Dawson.

The Las Campanas/Anglo-Australian Telescope Rich Cluster Survey - III. Spectroscopic studies of X-ray bright galaxy clusters at z similar to 0.1

We present the analysis of the spectroscopic and photometric catalogues of 11 X-ray luminous clusters at 0.07 < z < 0.16 from the Las Campanas/Anglo-Australian Telescope Rich Cluster Survey. Our spectroscopic data set consists of over 1600 galaxy cluster members, of which two-thirds are outside r(200). These spectra allow us to assign cluster membership using a detailed mass model and expand on our previous work on the cluster colour-magnitude relation ( CMR) where membership was inferred statistically. We confirm that the modal colours of galaxies on the CMR become progressively bluer with increasing radius d( B - R)/dr(p) = - 0.011 +/- 0.003 and with decreasing local galaxy density d( B - R)/dlog ( Sigma)= - 0.062 +/- 0.009. Interpreted as an age effect, we hypothesize that these trends in galaxy colour should be reflected in mean H delta equivalent width. We confirm that passive galaxies in the cluster increase in Hd line strength as dH delta/dr(p) = 0.35 +/- 0.06. Therefore, those galaxies in the cluster outskirts may have younger luminosity-weighted stellar populations; up to 3 Gyr younger than those in the cluster centre assuming d( B - R)/dt = 0.03 mag per Gyr. A variation of star formation rate, as measured by [ O II]lambda 3727 angstrom, with increasing local density of the environment is discernible and is shown to be in broad agreement with previous studies from the 2dF Galaxy Redshift Survey and the Sloan Digital Sky Survey. We divide our spectra into a variety of types based upon the MORPHs classification scheme. We find that clusters at z similar to 0.1 are less active than their higher-redshift analogues: about 60 per cent of the cluster galaxy population is non-star forming, with a further 20 per cent in the post-starburst class and 20 per cent in the currently active class, demonstrating that evolution is visible within the past 2 - 3 Gyr. We also investigate unusual populations of blue and very red non-star forming galaxies and we suggest that the former are likely to be the progenitors of galaxies which will lie on the CMR, while the colours of the latter possibly reflect dust reddening. We show that the cluster galaxies at large radii consist of both backsplash ones and those that are infalling to the cluster for the first time. We make a comparison to the field population at z similar to 0.1 and examine the broad differences between the two populations. Individually, the clusters show significant variation in their galaxy populations which we suggest reflects their recent infall histories.

Universal quantum computation with continuous-variable cluster states

We describe a generalization of the cluster-state model of quantum computation to continuous-variable systems, along with a proposal for an optical implementation using squeezed-light sources, linear optics, and homodyne detection. For universal quantum computation, a nonlinear element is required. This can be satisfied by adding to the toolbox any single-mode non-Gaussian measurement, while the initial cluster state itself remains Gaussian. Homodyne detection alone suffices to perform an arbitrary multimode Gaussian transformation via the cluster state. We also propose an experiment to demonstrate cluster-based error reduction when implementing Gaussian operations.

On the simultaneous use of clinical and microarray expression data in the cluster analysis of tissue samples

This paper considers a model-based approach to the clustering of tissue samples of a very large number of genes from microarray experiments. It is a nonstandard problem in parametric cluster analysis because the dimension of the feature space (the number of genes) is typically much greater than the number of tissues. Frequently in practice, there are also clinical data available on those cases on which the tissue samples have been obtained. Here we investigate how to use the clinical data in conjunction with the microarray gene expression data to cluster the tissue samples. We propose two mixture model-based approaches in which the number of components in the mixture model corresponds to the number of clusters to be imposed on the tissue samples. One approach specifies the components of the mixture model to be the conditional distributions of the microarray data given the clinical data with the mixing proportions also conditioned on the latter data. Another takes the components of the mixture model to represent the joint distributions of the clinical and microarray data. The approaches are demonstrated on some breast cancer data, as studied recently in van't Veer et al. (2002).

Cluster analysis of the p53 genetic regulatory network: topology and biology

We describe a network module detection approach which combines a rapid and robust clustering algorithm with an objective measure of the coherence of the modules identified. The approach is applied to the network of genetic regulatory interactions surrounding the tumor suppressor gene p53. This algorithm identifies ten clusters in the p53 network, which are visually coherent and biologically plausible.