Trading spaces : on the lore and limitations of latent semantic analysis

Two decades after its inception, Latent Semantic Analysis(LSA) has become part and parcel of every modern introduction to Information Retrieval. For any tool that matures so quickly, it is important to check its lore and limitations, or else stagnation will set in. We focus here on the three main aspects of LSA that are well accepted, and the gist of which can be summarized as follows: (1) that LSA recovers latent semantic factors underlying the document space, (2) that such can be accomplished through lossy compression of the document space by eliminating lexical noise, and (3) that the latter can best be achieved by Singular Value Decomposition. For each aspect we performed experiments analogous to those reported in the LSA literature and compared the evidence brought to bear in each case. On the negative side, we show that the above claims about LSA are much more limited than commonly believed. Even a simple example may show that LSA does not recover the optimal semantic factors as intended in the pedagogical example used in many LSA publications. Additionally, and remarkably deviating from LSA lore, LSA does not scale up well: the larger the document space, the more unlikely that LSA recovers an optimal set of semantic factors. On the positive side, we describe new algorithms to replace LSA (and more recent alternatives as pLSA, LDA, and kernel methods) by trading its l2 space for an l1 space, thereby guaranteeing an optimal set of semantic factors. These algorithms seem to salvage the spirit of LSA as we think it was initially conceived.

Reliability of clinical measurement for assessing spinal fusion: An experimental sheep study

Study Design. A sheep study designed to compare the accuracy of static radiographs, dynamic radiographs, and computed tomographic (CT) scans for the assessment of thoracolumbar facet joint fusion as determined by micro-CT scanning. Objective. To determine the accuracy and reliability of conventional imaging techniques in identifying the status of thoracolumbar (T13-L1) facet joint fusion in a sheep model. Summary of Background Data. Plain radiographs are commonly used to determine the integrity of surgical arthrodesis of the thoracolumbar spine. Many previous studies of fusion success have relied solely on postoperative assessment of plain radiographs, a technique lacking sensitivity for pseudarthrosis. CT may be a more reliable technique, but is less well characterized. Methods. Eleven adult sheep were randomized to either attempted arthrodesis using autogenous bone graft and internal fixation (n = 3) or intentional pseudarthrosis (IP) using oxidized cellulose and internal fixation (n = 8). After 6 months, facet joint fusion was assessed by independent observers, using (1) plain static radiography alone, (2) additional dynamic radiographs, and (3) additional reconstructed spiral CT imaging. These assessments were correlated with high-resolution micro-CT imaging to predict the utility of the conventional imaging techniques in the estimation of fusion success. Results. The capacity of plain radiography alone to correctly predict fusion or pseudarthrosis was 43% and was not improved using plain radiography and dynamic radiography with also a 43% accuracy. Adding assessment by reformatted CT imaging to the plain radiography techniques increased the capacity to predict fusion outcome to 86% correctly. The sensitivity, specificity, and accuracy of static radiography were 0.33, 0.55, and 0.43, respectively, those of dynamic radiography were 0.46, 0.40, and 0.43, respectively, and those of radiography plus CT were 0.88, 0.85, and 0.86, respectively. Conclusion. CT-based evaluation correlated most closely with high-resolution micro-CT imaging. Neither plain static nor dynamic radiographs were able to predict fusion outcome accurately. © 2012 Lippincott Williams & Wilkins.

A finite volume method for solving the two-sided time-space fractional advection-dispersion equation

The field of fractional differential equations provides a means for modelling transport processes within complex media which are governed by anomalous transport. Indeed, the application to anomalous transport has been a significant driving force behind the rapid growth and expansion of the literature in the field of fractional calculus. In this paper, we present a finite volume method to solve the time-space two-sided fractional advection dispersion equation on a one-dimensional domain. Such an equation allows modelling different flow regime impacts from either side. The finite volume formulation provides a natural way to handle fractional advection-dispersion equations written in conservative form. The novel spatial discretisation employs fractionally-shifted Gr¨unwald formulas to discretise the Riemann-Liouville fractional derivatives at control volume faces in terms of function values at the nodes, while the L1-algorithm is used to discretise the Caputo time fractional derivative. Results of numerical experiments are presented to demonstrate the effectiveness of the approach.

High level protein expression in plants through the use of a novel autonomously replicating geminivirus shuttle vector

We constructed a novel autonomously replicating gene expression shuttle vector, with the aim of developing a system for transiently expressing proteins at levels useful for commercial production of vaccines and other proteins in plants. The vector, pRIC, is based on the mild strain of the geminivirus Bean yellow dwarf virus (BeYDV-m) and is replicationally released into plant cells from a recombinant Agrobacterium tumefaciens Ti plasmid. pRIC differs from most other geminivirus-based vectors in that the BeYDV replication-associated elements were included in cis rather than from a co-transfected plasmid, while the BeYDV capsid protein (CP) and movement protein (MP) genes were replaced by an antigen encoding transgene expression cassette derived from the non-replicating A. tumefaciens vector, pTRAc. We tested vector efficacy in Nicotiana benthamiana by comparing transient cytoplasmic expression between pRIC and pTRAc constructs encoding either enhanced green fluorescent protein (EGFP) or the subunit vaccine antigens, human papillomavirus subtype 16 (HPV-16) major CP L1 and human immunodeficiency virus subtype C p24 antigen. The pRIC constructs were amplified in planta by up to two orders of magnitude by replication, while 50% more HPV-16 L1 and three- to seven-fold more EGFP and HIV-1 p24 were expressed from pRIC than from pTRAc. Vector replication was shown to be correlated with increased protein expression. We anticipate that this new high-yielding plant expression vector will contribute towards the development of a viable plant production platform for vaccine candidates and other pharmaceuticals. © 2009 Blackwell Publishing Ltd.

Insights into the role and function of L2, the minor capsid protein of papillomaviruses

Human papillomaviruses (HPV) are responsible for the most common human sexually transmitted viral infections, and high-risk types are responsible for causing cervical and other cancers. The minor capsid protein L2 of HPV plays important roles in virus entry into cells, localisation of viral components to the nucleus, in DNA binding, capsid formation and stability. It also elicits antibodies that are more cross-reactive between HPV types than does the major capsid protein L1, making it an attractive potential target for new-generation, more broadly protective subunit vaccines against HPV infections. However, its low abundance in natural capsids-12-72 molecules per 360 copies of L1-limits its immunogenicity. This review will explore the biological roles of the protein, and prospects for its use in new vaccines. © 2009 Springer-Verlag.

Comparison of cervical and blood T-cell responses to human papillomavirus-16 in women with human papillomavirus-associated cervical intraepithelial neoplasia

Human papillomaviruses (HPVs) are obligate epithelial pathogens and typically cause localized mucosal infections. We therefore hypothesized that T-cell responses to HPV antigens would be greater at sites of pathology than in the blood. Focusing on HPV-16 because of its association with cervical cancer, the magnitude of HPV-specific T-cell responses at the cervix was compared with those in the peripheral blood by intracellular cytokine staining following direct ex vivo stimulation with both virus-like particles assembled from the major capsid protein L1, and the major HPV oncoprotein, E7. We show that both CD4 + and CD8 + T cells from the cervix responded to the HPV-16 antigens and that interferon-γ (IFN-γ) production was HPV type-specific. Comparing HPV-specific T-cell IFN-γ responses at the cervix with those in the blood, we found that while CD4 + and CD8 + T-cell responses to L1 were significantly correlated between compartments (P = 0.02 and P = 0.05, respectively), IFN-γ responses in both T-cell subsets were significantly greater in magnitude at the cervix than in peripheral blood (P = 0.02 and P = 0.003, respectively). In contrast, both CD4 + and CD8 + T-cell IFN-γ responses to E7 were of similar magnitude in both compartments and CD8 + responses were significantly correlated between these distinct immunological compartments (P = 0.04). We therefore show that inflammatory T-cell responses against L1 (but not E7) demonstrate clear compartmental bias and the magnitude of these responses do reflect local viral replication but that correlation of HPV-specific responses between compartments indicates their linkage.

Strategies for the prevention of cervical cancer by human papillomavirus vaccination

As cervical cancer is causally associated with 14 high-risk types of human papillomavirus (HPV), a successful HPV vaccine will have a major impact on this disease. Although some persistent HPV infections progress to cervical cancer, host immunity is generally able to clear most HPV infections. Both cell-mediated and antibody responses have been implicated in influencing the susceptibility, persistence or clearance of genital HPV infection. There have been two clinical trials that show that vaccines based on virus-like particles (VLPs) made from the major capsid protein, L1, are able to type specifically protect against cervical intra-epithelial neoplasia and infection. However, there is no evidence that even a mixed VLP vaccine will protect against types not included in the vaccine, and a major challenge that remains is how to engineer protection across a broader spectrum of viruses. Strategies for production of HPV vaccines using different vaccine vectors and different production systems are also reviewed. © 2005 Elsevier Ltd. All rights reserved.

A deletion and point mutation study of the human papillomavirus type 16 major capsid gene

Recombinant human papillomavirus (HPV) virus-like particles (VLPs) made from the major capsid protein L1 are promising vaccine candidates for use as vaccines against genital and other HPV infections, and particularly against HPV-16. However, HPV-16 genotype variants have different binding affinities for neutralising mouse Mabs raised against HPV-16 L1 VLPs. This paper analyses, using a panel of well-characterised Mabs, the effects on the antigenicity of various C- and N-terminal deletants of HPV-16 L1 made in insect cells via recombinant baculovirus, of an A → T mutation at residue 266 (A266T), and of a C → G mutation at conserved position 428 (C428G). The effects of these changes on assembly of the variant L1s were studied by electron microscopy. Binding of Mab H16:E70 to A266T was reduced by almost half in comparison to wild type L1. Retention of the C-terminal region 428-483 was critical for the binding of conformation-specific Mabs (H16:V5, H16:E70, H16:U4 and H16:9A) whereas deletion of the nuclear localisation signal (NLS) or the C428G mutation or an N-terminal deletion (residues 2-9) did not affect the antigenicity. The N-terminal deletion resulted in a mixed population of 30 and 55 nm VLPs, which differs from the same construct expressed in Escherichia coli, whereas pentamer aggregates resulted from deletion of the 428-465 region or the C428G mutation. The results have implications both for considering use of single-genotype HPV vaccines, and for design of novel second-generation vaccines. © 2006 Elsevier B.V. All rights reserved.

Human papillomavirus vaccines in plants

Human papillomaviruses are the etiological agents of cervical cancer, one of the two most prevalent cancers in women in developing countries. Currently available prophylactic vaccines are based on the L1 major capsid protein, which forms virus-like particles when expressed in yeast and insect cell lines. Despite their recognized efficacy, there are significant shortcomings: the vaccines are expensive, include only two oncogenic virus types, are delivered via intramuscular injection and require a cold chain. Plant expression systems may provide ways of overcoming some of these problems, in particular the expense. In this article, we report recent promising advances in the production of prophylactic and therapeutic vaccines against human papillomavirus by expression of the relevant antigens in plants, and discuss future prospects for the use of such vaccines. © 2010 Expert Reviews Ltd.

Therapeutic immunisation of rabbits with cottontail rabbit papillomavirus (CRPV) virus-like particles (VLP) induces regression of established papillomas

There is overwhelming evidence that persistent infection with high-risk human papillomaviruses (HR-HPV) is the main risk factor for invasive cancer of the cervix. Due to this global public health burden, two prophylactic HPV L1 virus-like particles (VLP) vaccines have been developed. While these vaccines have demonstrated excellent type-specific prevention of infection by the homologous vaccine types (high and low risk HPV types), no data have been reported on the therapeutic effects in people already infected with the low-risk HPV type. In this study we explored whether regression of CRPV-induced papillomas could be achieved following immunisation of out-bred New Zealand White rabbits with CRPV VLPs. Rabbits immunised with CRPV VLPs had papillomas that were significantly smaller compared to the negative control rabbit group (P ≤ 0.05). This data demonstrates the therapeutic potential of PV VLPs in a well-understood animal model with potential important implications for human therapeutic vaccination for low-risk HPVs. © 2008 Govan et al; licensee BioMed Central Ltd.

Markers of cutaneous human papillomavirus infection in individuals with tumor-free skin, actinic keratoses, and squamous cell carcinoma

Separately, actinic keratosis (AK) and cutaneous squamous cell carcinoma (SCC) have been associated with cutaneous human papillomavirus (HPV) infections. To further explore the association between HPV infection and SCC development, we determined markers of cutaneous HPV infection within a single population in persons with precursor lesions (AK), cancerous lesions (SCC), and without. Serum and plucked eyebrow hairs were collected from 57 tumor-free controls, 126 AK, and 64 SCC cases. Presence of HPV L1 and E6 seroreactivity and viral DNA were determined for HPV types 5, 8, 15, 16, 20, 24, and 38. Significant positive associations with increasing severity of the lesions (controls, AK, and SCC, respectively) were observed for overall HPV L1 seropositivity (13%, 26%, and 37%) and for HPV8 (4%, 17%, and 30%). In parallel, the proportion of L1 seropositive individuals against multiple HPV types increased from 14% to 39% and 45%. The overall E6 seroreactivity, however, tended to decline with AK and SCC, especially for HPV8 (21%, 11%, and 2%). HPV DNA positivity was most prevalent in the AK cases (54%) compared with the SCC cases (44%) and the tumor-free controls (40%). Among all participants, there was a positive trend between overall HPV DNA positivity and L1 seropositivity, but not E6 seropositivity. Taken together, our data suggest that cutaneous HPV infections accompanied by detectable HPV DNA in eyebrow hairs and HPV L1 seropositivity, but not E6 seropositivity, are associated with an increased risk of AK and SCC.

Perceived social support and community adaptation in schizophrenia

Prompted by the continuing transition to community care, mental health nurses are considering the role of social support in community adaptation. This article demonstrates the importance of distinguishing between kinds of social support and presents findings from the first round data of a longitudinal study of community adaptation in 156 people with schizophrenia conducted in Brisbane, Australia. All clients were interviewed using the relevant subscales of the Diagnostic Interview Schedule to confirm a primary diagnosis of schizophrenia. The study set out to investigate the relationship between community adaptation and social support. Community adaptation was measured with the Brief Psychiatric Rating Scale (BPRS), the Life Skills Profile (LSP) and measures of dissatisfaction with life and problems in daily living developed by the authors. Social support was measured with the Arizona Social Support Interview Schedule (ASSIS). The BPRS and ASSIS were incorporated into a client interview conducted by trained interviewers. The LSP was completed on each client by an informal carer (parent, relative or friend) or a professional carer (case manager or other health professional) nominated by the client. Hierarchical regression analysis was used to examine the relationship between community adaptation and four sets of social support variables. Given the order in which variables were entered in regression equations, a set of perceived social support variables was found to account for the largest unique variance of four measures of community adaptation in 96 people with schizophrenia for whom complete data are available from the first round of the three-wave longitudinal study. A set of the subjective experiences of the clients accounted for the largest unique variance in measures of symptomatology, life skills, dissatisfaction with life, and problems in daily living. Sets of community support, household support and functional variables accounted for less variance. Implications for mental health nursing practice are considered.

On robust face recognition via sparse encoding : the good, the bad, and the ugly

In the field of face recognition, Sparse Representation (SR) has received considerable attention during the past few years. Most of the relevant literature focuses on holistic descriptors in closed-set identification applications. The underlying assumption in SR-based methods is that each class in the gallery has sufficient samples and the query lies on the subspace spanned by the gallery of the same class. Unfortunately, such assumption is easily violated in the more challenging face verification scenario, where an algorithm is required to determine if two faces (where one or both have not been seen before) belong to the same person. In this paper, we first discuss why previous attempts with SR might not be applicable to verification problems. We then propose an alternative approach to face verification via SR. Specifically, we propose to use explicit SR encoding on local image patches rather than the entire face. The obtained sparse signals are pooled via averaging to form multiple region descriptors, which are then concatenated to form an overall face descriptor. Due to the deliberate loss spatial relations within each region (caused by averaging), the resulting descriptor is robust to misalignment & various image deformations. Within the proposed framework, we evaluate several SR encoding techniques: l1-minimisation, Sparse Autoencoder Neural Network (SANN), and an implicit probabilistic technique based on Gaussian Mixture Models. Thorough experiments on AR, FERET, exYaleB, BANCA and ChokePoint datasets show that the proposed local SR approach obtains considerably better and more robust performance than several previous state-of-the-art holistic SR methods, in both verification and closed-set identification problems. The experiments also show that l1-minimisation based encoding has a considerably higher computational than the other techniques, but leads to higher recognition rates.

Generalized-positioning for mixed-frequency of mixed-GNSS and its preliminary applications

Modernized GPS and GLONASS, together with new GNSS systems, BeiDou and Galileo, offer code and phase ranging signals in three or more carriers. Traditionally, dual-frequency code and/or phase GPS measurements are linearly combined to eliminate effects of ionosphere delays in various positioning and analysis. This typical treatment method has imitations in processing signals at three or more frequencies from more than one system and can be hardly adapted itself to cope with the booming of various receivers with a broad variety of singles. In this contribution, a generalized-positioning model that the navigation system independent and the carrier number unrelated is promoted, which is suitable for both single- and multi-sites data processing. For the synchronization of different signals, uncalibrated signal delays (USD) are more generally defined to compensate the signal specific offsets in code and phase signals respectively. In addition, the ionospheric delays are included in the parameterization with an elaborate consideration. Based on the analysis of the algebraic structures, this generalized-positioning model is further refined with a set of proper constrains to regularize the datum deficiency of the observation equation system. With this new model, uncalibrated signal delays (USD) and ionospheric delays are derived for both GPS and BeiDou with a large dada set. Numerical results demonstrate that, with a limited number of stations, the uncalibrated code delays (UCD) are determinate to a precision of about 0.1 ns for GPS and 0.4 ns for BeiDou signals, while the uncalibrated phase delays (UPD) for L1 and L2 are generated with 37 stations evenly distributed in China for GPS with a consistency of about 0.3 cycle. Extra experiments concerning the performance of this novel model in point positioning with mixed-frequencies of mixed-constellations is analyzed, in which the USD parameters are fixed with our generated values. The results are evaluated in terms of both positioning accuracy and convergence time.

Secretome and degradome profiling shows that Kallikrein-related peptidases 4, 5, 6, and 7 induce TGFβ-1 signaling in ovarian cancer cells

Kallikrein-related peptidases, in particular KLK4, 5, 6 and 7 (4-7), often have elevated expression levels in ovarian cancer. In OV-MZ-6 ovarian cancer cells, combined expression of KLK4-7 reduces cell adhesion and increases cell invasion and resistance to paclitaxel. The present work investigates how KLK4-7 shape the secreted proteome ("secretome") and proteolytic profile ("degradome") of ovarian cancer cells. The secretome comparison consistently identified >900 proteins in three replicate analyses. Expression of KLK4-7 predominantly affected the abundance of proteins involved in cell-cell communication. Among others, this includes increased levels of transforming growth factor β-1 (TGFβ-1). KLK4-7 co-transfected OV-MZ-6 cells share prominent features of elevated TGFβ-1 signaling, including increased abundance of neural cell adhesion molecule L1 (L1CAM). Augmented levels of TGFβ-1 and L1CAM upon expression of KLK4-7 were corroborated in vivo by an ovarian cancer xenograft model. The degradomic analysis showed that KLK4-7 expression mostly affected cleavage sites C-terminal to arginine, corresponding to the preference of kallikreins 4, 5 and 6. Putative kallikrein substrates include chemokines, such as growth differentiation factor 15 (GDF 15) and macrophage migration inhibitory factor (MIF). Proteolytic maturation of TGFβ-1 was also elevated. KLK4-7 have a pronounced, yet non-degrading impact on the secreted proteome, with a strong association between these proteases and TGFβ-1 signaling in tumor biology. © 2013 Federation of European Biochemical Societies.