Simultaneous feature selection and feature weighting using Hybrid Tabu Search/K-nearest neighbor classifier

Feature selection and feature weighting are useful techniques for improving the classification accuracy of K-nearest-neighbor (K-NN) rule. The term feature selection refers to algorithms that select the best subset of the input feature set. In feature weighting, each feature is multiplied by a weight value proportional to the ability of the feature to distinguish pattern classes. In this paper, a novel hybrid approach is proposed for simultaneous feature selection and feature weighting of K-NN rule based on Tabu Search (TS) heuristic. The proposed TS heuristic in combination with K-NN classifier is compared with several classifiers on various available data sets. The results have indicated a significant improvement in the performance in classification accuracy. The proposed TS heuristic is also compared with various feature selection algorithms. Experiments performed revealed that the proposed hybrid TS heuristic is superior to both simple TS and sequential search algorithms. We also present results for the classification of prostate cancer using multispectral images, an important problem in biomedicine.

The endogenous granulocyte colony-stimulating factor response following autologous peripheral blood stem cell transplantation is inpaired in patients with myeloma

Granulocyte colony-stimulating factor (G-CSF) levels were studied in 23 patients (10 myeloma, 13 relapsed Hodgkin's disease, non-Hodgkin's lymphoma or germ cell tumours), post autologous peripheral blood stem cell transplantation (PBSCT). The two groups had similar previous chemotherapy and numbers of CD34+ cells transplanted. All patients received G-CSF by injection starting 8 d post transplantation. Twenty out of 23 patients showed raised endogenous levels of G-CSF before cytokine administration. Myeloma patients showed significantly lower levels of endogenous G-CSF than the other patients (0.767 versus 3.262 ng/ml, P <0.05). Further rises in G-CSF levels were seen following the administration of exogenous G-CSF which then fell, despite ongoing administration of G-CSF, as neutrophil recovery occurred.

Hypoxia selects for androgen independent LNCaP cells with a more malignant geno- and phenotype.

Hypoxia confers resistance to common cancer therapies, however, it has also has been shown to result in genetic alterations which may allow a survival advantage and increase the tumorigenic properties of cancer cells. Additionally, it may exert a selection pressure, allowing expansion of tumor cells with a more aggressive phenotype. To further assess the role of hypoxia in malignant progression in prostate cancer we exposed human androgen dependent prostate cancer cells (LNCaP) to cycles of chronic hypoxia and isolated a subline, LNCaP-H1. This article describes the partial characterization of this cell line. The LNCaP-H1 subline showed altered growth characteristics and exhibited androgen independent growth both in vitro and in vivo. Furthermore, these cells were resistant to mitochondrial-mediated apoptosis, probably since the endogenous levels of Bax was lower and Bcl-2 higher than in the parental LNCaP cells. Microarray analysis revealed that a complex array of pathways had differential gene expression between the 2 cell lines, with LNCaP-H1 cells exhibiting a genetic profile which suggests that they may be more likely metastasize to distant organs, especially bone. This was supported by an in vitro invasion assay, and an in vivo metastasis study. This study shows that hypoxia can select for androgen independent prostate cancer cells which have a survival advantage and are more likely to invade and metastasize.

Model selection approaches for non-linear system identification: a review

The identification of non-linear systems using only observed finite datasets has become a mature research area over the last two decades. A class of linear-in-the-parameter models with universal approximation capabilities have been intensively studied and widely used due to the availability of many linear-learning algorithms and their inherent convergence conditions. This article presents a systematic overview of basic research on model selection approaches for linear-in-the-parameter models. One of the fundamental problems in non-linear system identification is to find the minimal model with the best model generalisation performance from observational data only. The important concepts in achieving good model generalisation used in various non-linear system-identification algorithms are first reviewed, including Bayesian parameter regularisation and models selective criteria based on the cross validation and experimental design. A significant advance in machine learning has been the development of the support vector machine as a means for identifying kernel models based on the structural risk minimisation principle. The developments on the convex optimisation-based model construction algorithms including the support vector regression algorithms are outlined. Input selection algorithms and on-line system identification algorithms are also included in this review. Finally, some industrial applications of non-linear models are discussed.

Gene selection by cooperative competition clustering

Clustering analysis of data from DNA microarray hybridization studies is an essential task for identifying biologically relevant groups of genes. Attribute cluster algorithm (ACA) has provided an attractive way to group and select meaningful genes. However, ACA needs much prior knowledge about the genes to set the number of clusters. In practical applications, if the number of clusters is misspecified, the performance of the ACA will deteriorate rapidly. In fact, it is a very demanding to do that because of our little knowledge. We propose the Cooperative Competition Cluster Algorithm (CCCA) in this paper. In the algorithm, we assume that both cooperation and competition exist simultaneously between clusters in the process of clustering. By using this principle of Cooperative Competition, the number of clusters can be found in the process of clustering. Experimental results on a synthetic and gene expression data are demonstrated. The results show that CCCA can choose the number of clusters automatically and get excellent performance with respect to other competing methods.

Selection with melphalan or paclitaxel (Taxol) yields variants with different patterns of multidrug resistance, integrin expression and in vitro invasiveness

A melphalan-resistant variant (Roswell Park Memorial Institute (RPMI)-2650M1) and a paclitaxel-resistant variant (RPMI-1650Tx) of the drug-sensitive human nasal carcinoma cell line, RPMI-2650. were established. The multidrug resistance (MDR) phenotype in the RPMI-2650Tx appeared to be P-glycoprotein (PgP)-mediated. Overexpression of multidrug resistant protein (MRP) family members was observed in the RPMI-2650M1 cells, which were also much more invasive in vitro than the parental cell line or the paclitaxel-resistant variant. Increased expression of alpha (2), alpha (5), alpha (6), beta (1) and beta (4) integrin subunits, decreased expression of alpha (4) integrin subunit, stronger adhesion to collagen type IV, laminin, fibronectin and matrigel, increased expression of MMP-2 and MMP-9 and significant motility compared with the parental cells were observed, along with a high invasiveness in the RPMI-7650M1 cells. Decreased expression of the alpha (2) integrin subunit, decreased attachment to collagen type IV, absence of cytokeratin 18 expression, no detectable expression of gelatin-degrading proteases and poor motility may be associated with the non-invasiveness of the RPMI-2650Tx variant. These results suggest that melphalan exposure can result in not only a MDR phenotype. but could also make cancer cells more invasive, whereas paclitaxel exposure resulted in MDR without increasing the in vitro invasiveness in the RPMI-2650 cells. (C) 2001 Elsevier Science Ltd. All rights reserved.

Kir2.2 inward rectifier potassium channels are inhibited by an endogenous factor in Xenopus oocytes independently from the action of a mitochondrial uncoupler.

We previously showed inhibition of Kir2 inward rectifier K+ channels expressed in Xenopus oocytes by the mitochondrial agents carbonyl cyanide p-trifluoromethoxyphenylhydrazone (FCCP) and sodium azide. Mutagenesis studies suggested that FCCP may act via phosphatidylinositol 4,5-bisphosphate (PIP2) depletion. This mechanism could be reversible in intact cells but not in excised membrane patches which preclude PIP2 regeneration. This prediction was tested by investigating the reversibility of the inhibition of Kir2.2 by FCCP in intact cells and excised patches. We also investigated the effect of FCCP on Kir2.2 expressed in human embryonic kidney (HEK) cells. Kir2.2 current, expressed in Xenopus oocytes, increased in inside-out patches from FCCP-treated and untreated oocytes. The fraction of total current that increased was 0.79?±?0.05 in control and 0.89?±?0.03 in 10 µM FCCP-treated (P?>?.05). Following “run-up,” Kir2.2 current was re-inhibited by “cramming” inside-out patches into oocytes. Therefore, run-up reflected not reversal of inhibition by FCCP, but washout of an endogenous inhibitor. Kir2.2 current recovered in intact oocytes within 26.5 h of FCCP removal. Injection of oocytes with 0.1 U apyrase completely depleted ATP (P?<?.001) but did not inhibit Kir2.2 and inhibited Kir2.1 by 35% (P?<?.05). FCCP only partially reduced [ATP] (P?<?.001), despite inhibiting Kir2.2 by 75% (P?<?.01) but not Kir2.1. FCCP inhibited Kir2.2 expressed in HEK cells. The recovery of Kir2.2 from inhibition by FCCP requires intracellular components, but direct depletion of ATP does not reproduce the differential inhibitory effect of FCCP. Inhibition of Kir2.2 by FCCP is not unique to Xenopus oocytes. J. Cell. Physiol. 219: 8–13, 2009. © 2008 Wiley-Liss, Inc.