The application of adaptive partitioned random search in feature selection problem

Feature selection is one of important and frequently used techniques in data preprocessing. It can improve the efficiency and the effectiveness of data mining by reducing the dimensions of feature space and removing the irrelevant and redundant information. Feature selection can be viewed as a global optimization problem of finding a minimum set of M relevant features that describes the dataset as well as the original N attributes. In this paper, we apply the adaptive partitioned random search strategy into our feature selection algorithm. Under this search strategy, the partition structure and evaluation function is proposed for feature selection problem. This algorithm ensures the global optimal solution in theory and avoids complete randomness in search direction. The good property of our algorithm is shown through the theoretical analysis.

Elastic moduli of model random three-dimensional closed-cell cellular solids

Most cellular solids are random materials, while practically all theoretical structure-property results are for periodic models. To be able to generate theoretical results for random models, the finite element method (FEM) was used to study the elastic properties of solids with a closed-cell cellular structure. We have computed the density (rho) and microstructure dependence of the Young's modulus (E) and Poisson's ratio (PR) for several different isotropic random models based on Voronoi tessellations and level-cut Gaussian random fields. The effect of partially open cells is also considered. The results, which are best described by a power law E infinity rho (n) (1<n<2), show the influence of randomness and isotropy on the properties of closed-cell cellular materials, and are found to be in good agreement with experimental data. (C) 2001 Acta Materialia Inc. Published by Elsevier Science Ltd. All rights reserved.

Long-term survivor mixture model with random effects: application to a multi-centre clinical trial of carcinoma

A mixture model incorporating long-term survivors has been adopted in the field of biostatistics where some individuals may never experience the failure event under study. The surviving fractions may be considered as cured. In most applications, the survival times are assumed to be independent. However, when the survival data are obtained from a multi-centre clinical trial, it is conceived that the environ mental conditions and facilities shared within clinic affects the proportion cured as well as the failure risk for the uncured individuals. It necessitates a long-term survivor mixture model with random effects. In this paper, the long-term survivor mixture model is extended for the analysis of multivariate failure time data using the generalized linear mixed model (GLMM) approach. The proposed model is applied to analyse a numerical data set from a multi-centre clinical trial of carcinoma as an illustration. Some simulation experiments are performed to assess the applicability of the model based on the average biases of the estimates formed. Copyright (C) 2001 John Wiley & Sons, Ltd.

Methylation of O-6-methylguanine DNA methyltransferase characterizes a subset of colorectal cancer with low-level DNA microsatellite instability

The significance of low-level DNA microsatellite instability (MSI-L) is not well understood. K-ras mutation is associated with MSI-L colorectal cancer and with the silencing of the DNA repair gene O-6-methylguanine DNA methyltransferase (MGMT) by methylation of its promoter region. MGMT methylation was studied in sporadic colorectal cancers stratified as DNA microsatellite instability-high (n = 23), MSI-L (n = 44), and microsatellite-stable (n = 23). Methylation-specific PCR was used to detect MGMT-promoter hypermethylation in 3 of 23 (13%) microsatellite instability-high, in 28 of 44 (64%) MSI-L, and in 6 of 23 (26%) microsatellite-stable cancers (P = 0.0001). K-ras was mutated in 20 of 29 (69%) methylated MSI-L cancers and in 2 of 15 (13%) unmethylated MSI-L cancers (P = 0.001), indicating a relationship between MGMT-methylation and mutation of K-ras. Loss of nuclear expression of MGMT was demonstrated immunohistochemically in 23 of 31 (74%) cancers with methylated MGMT and in 10 of 49 (20%) cancers with nonmethylated MGMT (P < 0.0001). Loss of expression of MGMT was also demonstrated in 9 of 31 serrated polyps. Silencing of MGMT may predispose to mutation by overwhelming the DNA mismatch repair system and occurs with greatest frequency in MSI-L colorectal cancers.

HLA class II antigens positively and negatively associated with hepatosplenic schistosomiasis in a Chinese population

To identify possible associations between host genetic factors and the onset of liver fibrosis following Schistosoma japonicum infection, the major histocompatibility class II alleles of 84 individuals living on an island (Jishan) endemic for schistosomiasis japonica in the Poyang Lake Region of Southern China were determined. Forty patients exhibiting advanced schistosomiasis, characterised by extensive liver fibrosis, and 44 age and sex-matched control subjects were assessed for the class II haplotypes HLA-DRBI and HLA-DQB1. Two HLA-DRB1 alleles, HLA-DRB1*0901 (P = 0.012) and *1302 (P = 0.039), and two HLA-DQB1 alleles, HLA-DQB1*0303 (P = 0.012) and *0609 (P = 0.037), were found to be significantly associated with susceptibility to fibrosis. These associated DRB1 and DQB1 alleles are in very strong linkage disequilibrium, with DRB1*0901-DQB1*0303 and DRB1*1302-DQB1*0609 found as: common haplotypes in this population. In contrast, the alleles HLA-DRB1*1501 (P = 0.025) and HLA-DQB 1*0601 (P = 0.022) were found to be associated with resistance to hepatosplenic disease. Moreover, the alleles DQB1*0303 and DRB1*0901 did not increase susceptibility in the presence of DQB1*0601, indicating that DQB1*0601 is dominant over DQB1*0303 and DRB1*0901. The study has thus identified both positive and negative associations between HLA class II alleles and the risk of individuals developing moderate to severe liver fibrosis following schistosome infection. (C) 2001 Australian Society for Parasitology Inc. Published by Elsevier Science Ltd. All rights reserved.

Identification of an apolipoprotein(e) variant associated with type III hyperlipoproteinaemia in an indigenous Australian

As a result of testing for lipid and apolipoprotein(e) (apo E) phenotype status of an indigenous Australian community, an apo E variant associated with type III hyperlipoproteinaemia has been identified. Apo E phenotype was determined by analysis of VLDL by isoelectric focusing, and genotype on DNA amplified by polymerase chain reaction, using two different restriction enzyme isotyping assays. Phenotypes and genotypes were discordant in samples from two subjects and an abnormal-sized restriction fragment was also observed in their genotyping gel patterns. DNA sequencing studies revealed this was due to a single nucleotide deletion. 3817delC, at amino acid 136 on apo E. This resulted in a new reading frame and the premature termination of the apo E protein due to a stop codon (TGA) at nucleotide 4105. The variant apo E null allele showed a recessive mode of inheritance and, in combination with the E2 allele, resulted in the type III hyperlipoproteinaemic phenotype but when inherited with the E4 allele had no marked effect on plasma lipids.

Ultrastructure and small-subunit ribosomal DNA sequence of Henneguya lesteri n. sp (Myxosporea), a parasite of sand whiting Sillago analis (Sillaginidae) from the coast of Queensland, Australia

Henneguya lesteri n. sp, (Myxosporea) is described from sand whiting, Sillago analis, from the southern Queensland coast of Australia. H. lesteri displays a preference for the pseudobranchs and is typically positioned along the afferent blood vessels, displacing the adjoining lamellae and disrupting their normal array, The plasmodia appeared as whitish-hyaline, elliptical cysts (mean dimensions 230 x 410 mum) attached to the oral mucosa lining of the hyoid arch on the inner surface of the operculum. Infections of the gills were also found, in which the plasmodia were spherical, averaged 240 x 240 mum in size and were located on the inner hemibranch margin. The parasites lodged in the gill filament crypts and generated a mild hyperplastic response of the branchial epithelium, In histological sections, the plasmodium wall and adjoining ectoplasm appeared as a finely granulated, weakly eosinophilic layer, Ultrastructurally, this section of the host-parasite interface contained an intricate complex of pinocytotic channels. H. lesteri is polysporic, disporoblastic and pansporoblast forming. Sporogenesis is asynchronous, with the earliest developmental stages aligned predominantly along the plasmodium periphery, and maturing sporoblasts and spores toward the center. Ultrastructural details of sporoblast and spore development are in agreement with previously described myxosporeans. The mature spore is drop-shaped, length (mean) 9.1 mum, width 4.7 mum, thickness 2.5 mum, and comprises 2 polar capsules positioned closely together, a binucleated sporoplasm and a caudal process of 12.6 mum. The polar capsules are elongated, 3.2 x 1.6 mum, with 4 turns of the polar filament. Mean length of the everted filament is 23.2 mum, Few studies have analyzed the 18S gene-of marine Myxosporea. In fact, H. lesteri is the first marine species of Henneguya to be characterized at the molecular level: we determined 1966 bp of the small-subunit (18S) rDNA, The results indicated that differences between this and the hitherto studied freshwater Henneguya species are greater than differences among the freshwater Henneguya species.

Efficiency of delivery of DNA to cells by bovine papillomavirus type-1 L1/L2 pseudovirions

To investigate the efficiency of encapsidation of plasmid by papillomavirus virus-like particles (PV VLPs), and the infectivity of the resultant PV pseudovirions, Cos-1 cells were transfected with an 8-kb plasmid incorporating a green fluorescent protein (GFP) reporter gene (pGSV), and infected with bovine PV (BPV-1) L1/L2 recombinant vaccinia virus to produce BPV1 pseudovirions. Approximately 1 in 1.5x10(4) of dense (1.35 g/ml) PV pseudovirions and 0.3 in 10(4) Of less-dense (1.29 g/ml) pseudovirions packaged an intact pGSV plasmid. The majority (>75%) of packaged plasmids contained deletions, and the deletions affected all tested genes. After exposure of Cos-1 cells to BPV-1 pseudovirions at an MOI of 40,000:1, 6% of cells expressed GFP giving a calculated efficiency of delivery of the pGSV plasmid, by pseudovirions which had packaged an intact plasmid, of approximately 5%. Plasmid delivery was not effected by purified pGSV plasmid, was blocked by antiserum against BPV-1, and was not blocked by DNase treatment of pseudovirions, confirming that delivery was mediated by DNA within the pseudovirion. We conclude that a major limitation to the use of PV pseudovirions as a gene delivery system is that intact plasmid DNA is not efficiently selected for packaging by VLPs in cell-based pseudovirions production systems.

X-linked neonatal diabetes mellitus, enteropathy and endocrinopathy syndrome is the human equivalent of mouse scurfy

To determine whether human X-linked neonatal diabetes mellitus, enteropathy and endocrinopathy syndrome (IPEX; MIM 304930) is the genetic equivalent of the scurfy (sf) mouse, we sequenced the human ortholog (FOXP3) of the gene mutated in scurfy mice (Foxp3), in IPEX patients. We found four non-polymorphic mutations. Each mutation affects the forkhead/winged-helix domain of the scurfin protein, indicating that the mutations may disrupt critical DNA interactions.

Application of DNA microarrays to study human alcoholism

An emerging idea is that long-term alcohol abuse results in changes in gene expression in the brain and that these changes are responsible at least partly for alcohol tolerance, dependence and neurotoxicity, The overall goal of our research is to identify genes which are differentia[ly expressed in the brains of well-characterized human alcoholics as compared with non-alcoholics. This should identify as-yet-unknown alcohol-responsive genes, and may well confirm changes in the expression of genes which have been delineated in animal models of alcohol abuse. Cases were carefully selected and samples pooled on the basis of relevant criteria; differential expression was monitored by microarray hybridization. The inherent diversity of human alcoholics can be exploited to identify genes associated with specific pathological processes, as well as to assess the effects of concomitant disease, severity of brain damage, drinking behavior, and factors such as gender and smoking history. initial results show selective changes in gene expression in alcoholics; of particular importance is a coordinated reduction in genes coding for myelin components, Copyright (C) 2001 National Science Council, ROC and S. Karger AG, Basel.

DNA adducts as a biomarker of polycyclic aromatic hydrocarbon exposure in aquatic organisms: relationship to carcinogenicity

The use of DNA adduct measurement as a biomarker of exposure to polycyclic aromatic hydrocarbons (PAHs) is now well established in ecotoxicology. In particular, DNA adduct levels in aquatic organisms has been found to produce a better correlation with PAH exposure than PAH concentrations in organisms. DNA adducts levels are most commonly determined using the P-32-postlabelling assay which measures total aromatic adducts. The relationship between relative DNA adduct formation and carcinogenicity has been investigated for a number of carcinogenic and non-carcinogenic PAHs using an in vitro system. Our results demonstrate that relatively high levels of DNA adducts can be produced by some non-carcinogenic PAHs, while other non-carcinogenic compounds do not produce detectable adducts. In addition, it has been shown that all carcinogenic PAHs investigated produce DNAadducts and that a relationship exists between relative adduct formation and carcinogenic potency. An investigation of adduct levels in fish liver and crustacean hepatopancreas in Oxley Ck, Brisbane has shown that higher than expected DNA adduct levels were correlated with the presence of carcinogenic and noncarcinogenic PAHs with high relative adduct forming potential.