China
Africa
Explore chapters and articles related to this topic
Angiogenesis and Roles of Adhesion Molecules in Psoriatic Disease
Published in Siba P. Raychaudhuri, Smriti K. Raychaudhuri, Debasis Bagchi, Psoriasis and Psoriatic Arthritis, 2017
Asmita Hazra, Saptarshi Mandal
TSP, a matricellular protein, may regulate angiogenesis in dose-dependent and context-dependent ways and among some of the most pleiotropic regulators of angiogenesis. As early as 1994, Nicosia and Tuszynski (1994) had observed concentration-dependent microvascular outgrowth from aortic ring explants in collagen and fibrin matrices containing TSP1. Wound healing was also paradoxically delayed in TSP1-null and TSP1 and 2 double-null mice and was accompanied by a reduction in blood vessels and inflammatory cells, as opposed to TSP2-null, where wounds heal faster than normal, which was initially thought to be due to reduced macrophage chemotaxis and decreased latent TGFβ activation. TSP1 receptors CD36 and β1 integrin associate with the VEGFR2. The coclustering of receptors that regulate angiogenesis may provide the endothelial cell with a platform for the integration of positive and negative signals. Such clusters have sometimes been considered modular signallosomes called “angosomes.” TSP1 can be directly pro-angiogenic in some contexts, for example, when working with syndecan 4.
Analysis of DNA Microarrays
Published in John Crowley, Antje Hoering, Handbook of Statisticsin Clinical Oncology, 2012
Shigeyuki Matsui, Hisashi Noma
Biclustering or two-way clustering, simultaneously, cluster genes and samples with the goal of identifying groups of genes involved in multiple biological activities in subsets of samples. A simple two-way clustering could be found by reordering the genes and samples after independently clustering them, such as available in the Eisen software (Eisen et al. 1998). More complex methods include coupled clustering (Getz et al. 2000), block clustering (Alon et al. 1999), the plaid model (Lazzeroni and Owen 2002), and Bayesian biclustering (Sheng et al. 2003). See Madeira and Oliveira (2004) for a review of biclustering algorithms.
Mapping knowledge structures and theme trends of atopic dermatitis: a co-word biclustering and quantitative analysis of the publication between 2015 and 2019
Published in Journal of Dermatological Treatment, 2022
Zhenzhen Mu, Yue Zhang, Lin Li, Xiuping Han
This work is based on a bibliometric analysis to reveal the current research trends in the realm of AD. Bibliometrics refers to a set of a particular method that can be carried out for deciphering and quantitatively analyzing the hot spots in the available literature. The most commonly employed method is Co-word analysis which has the potential to estimate the correlation of two words of professional significance in pertinent publications. Biclustering analysis is done in an attempt to cluster columns and lines simultaneously and conduct a partial analysis from the professional words that have been extracted. Moreover, the relation and evolutionary trends of themes were analyzed making use of SNA and strategic diagram. SNA was also employed to furnish a visual knowledge structure of AD alongside its clinical manifestations, pathogenesis, etiology, and treatment. In this manner, in addition to authors, journals and countries, the work also examines the attributes, theme trends, knowledge structure, and internal relationship of the published literature in the AD area between 2015 and 2019. A set of guidelines for future AD studies, from a bibliometric point of view, are summarized herein for the scientific community including researchers, medical educators, and clinical doctors.
Polygenic and Network-based studies in risk identification and demystification of cancer
Published in Expert Review of Molecular Diagnostics, 2022
Christopher El Hadi, Georges Ayoub, Yara Bachir, Michèle Haykal, Nadine Jalkh, Hampig Raphael Kourie
Van Dam et al. showed that computational clustering can be used to group genes with similar expression profiles across multiple samples, which should simplify cancer modeling [52]. Modules are the obvious outcome and can often be associated with biological processes and phenotypes [53–55]. The most widely used clustering method is Weighted Gene Correlation Network Analysis (WGCNA), which constructs gene co-expression modules using hierarchical clustering after correlating genes using their Microarray or RNAseq expression data [53,54]. Gene co-expression networks do not depend on prior gene information, avoid biologically incorrect assumptions about the independence of gene expression levels, and relieve researchers of the problems of multiple statistical testing [39]. For example, using these module-based inferences, it was found that humans and mice share fundamental transcriptional programs during early development that diverge at later stages [55]. Other methods, such as Generalized Single Value Decomposition (GSVD) and biclustering, identify modules and other properties of graphs that may be useful in cancer research. These methods take into account the heterogeneity of cancer and its evolution over time [52].
Assigning scores for ordered categorical responses
Published in Journal of Applied Statistics, 2020
Daniel Fernández, Ivy Liu, Roy Costilla, Peter Yongqi Gu
Let the PB data be represented by an Y where i and column j in Y . In this section, we give a brief description on the model with row (question) clustering (see [16] for a general biclustering model) as follows: 8), i is classified in the row cluster r and R row-clusters is unknown. Further, we define a priori row membership probabilities. In that way, we may deduce the number of parameters. Based on Model (8), the response probabilities are as follows: R clusters: a priori row membership probability of row i, and 9). This likelihood is constructed under the local independent assumption, that is, given parameters 16] developed a model fitting procedure using the EM algorithm [14,38]. which is commonly used in the case of the estimation of the parameters for a finite mixture-density model with incomplete data.