Your browser doesn't support javascript.
Mostrar: 20 | 50 | 100
Resultados 1 - 6 de 6
Mais filtros

Base de dados
Intervalo de ano de publicação
Stat Appl Genet Mol Biol ; 18(5)2019 09 17.
Artigo em Inglês | MEDLINE | ID: mdl-31525158


We propose a new bi-level feature selection method for high dimensional accelerated failure time models by formulating the models to a single index model. The method yields sparse solutions at both the group and individual feature levels along with an expedient algorithm, which is computationally efficient and easily implemented. We analyze a genomic dataset for an illustration, and present a simulation study to show the finite sample performance of the proposed method.

Stat Med ; 38(23): 4574-4582, 2019 Oct 15.
Artigo em Inglês | MEDLINE | ID: mdl-31304613


Feature selection is an important initial step of exploratory analysis in biomedical studies. Its main objective is to eliminate the covariates that are uncorrelated with the outcome. For highly correlated covariates, traditional feature selection methods, such as the Lasso, tend to select one of them and eliminate the others, although some of the eliminated ones are still scientifically valuable. To alleviate this drawback, we propose a feature selection method based on covariate space decomposition, referred herein as the "Decomposition Feature Selection" (DFS), and show that this method can lead to scientifically meaningful results in studies with correlated high dimensional data. The DFS consists of two steps: (i) decomposing the covariate space into disjoint subsets such that each of the subsets contains only uncorrelated covariates and (ii) identifying significant predictors by traditional feature selection within each covariate subset. We demonstrate through simulation studies that the DFS has superior practical performance over the Lasso type methods when multiple highly correlated covariates need to be retained. Application of the DFS is demonstrated through a study of bipolar disorders with correlated biomarkers.

ACS Appl Mater Interfaces ; 10(47): 40951-40957, 2018 Nov 28.
Artigo em Inglês | MEDLINE | ID: mdl-30338983


Interface engineering is an important approach leading to multifunctional artificial materials. Although most of the previous works focused on the effects of the rotation/tilting of interfacial oxygen octahedron on perovskite multilayers, here, we report on a new kind of lattice distortion characterized by an off-center shift of the Mn ions within the MnO6 oxygen octahedra at the interfaces of La1- xSr xCoO3/La2/3Sr1/3MnO3/La1- xSr xCoO3/LaAlO3 trilayers ( x = 0-1/3), which drives the initially perpendicularly aligned magnetic axis of the La2/3Sr1/3MnO3 (LSMO) film toward the in-plane direction, though the film is in a strongly compressive state. It is further found that the magnetic anisotropy considerably depends on the content of Sr in La1- xSr xCoO3, enhancing as x decreases. The maximal anisotropy constant at 10 K is +2.5 × 106 erg/cm3 for the trilayers with x = 0, whereas it is -1.5 × 105 erg/cm3 for a bare LSMO film on LaAlO3. On the basis of the analysis of X-ray absorption spectroscopy and the results of density functional theory calculations, we found that the off-center displacement of the Mn ions has caused a strong orbital reconstruction at interfaces, resulting in the anomalous spin orientation against magnetoelastic coupling.

PLoS One ; 12(6): e0180161, 2017.
Artigo em Inglês | MEDLINE | ID: mdl-28662145


Acute bacterial meningitis is still considered one of the most dangerous infectious diseases in children. To investigate the prevalence and antibiotic resistance profiles of cerebrospinal fluid (CSF) pathogens in children with acute bacterial meningitis in Southwest China, CSF samples from 179 meningitis patients (3 days to 12 years old) with positive culture results were collected from 2012 to 2015. Isolated pathogens were identified using the Vitek-32 system. Gram stain results were used to guide subcultures and susceptibility testing. The antimicrobial susceptibility of isolates was determined using the disc diffusion method. Of the isolates, 50.8% were Gram-positive bacteria, and 49.2% were Gram-negative bacteria. The most prevalent pathogens were E. coli (28.5%), Streptococcus pneumoniae (17.8%), Staphylococcus epidermidis (10.0%), Haemophilus influenzae type b (9.5%), and group B streptococcus (7.2%). In young infants aged ≤3 months, E. coli was the organism most frequently isolated from CSF (39/76; 51.3%), followed by group B streptococcus (13/76; 17.1%) and Streptococcus pneumoniae (8/76; 10.5%). However, in young infants aged >3 months, the most frequently isolated organism was Streptococcus pneumoniae (24/103; 23.3%), followed by Staphylococcus epidermidis (18/103; 17.5%) and Haemophilus influenzae type b (16/103; 15.5%). Antimicrobial susceptibility tests indicated that for E. coli isolates, the susceptibility rates to aminoglycosides ranged from 56.8% to 100.0%, among them, amikacin was identified as the most effective against E. coli. As for cephalosporins, the susceptibility rates ranged from 29.4% to 78.4%, and cefoxitin was identified as the most effective cephalosporin. In addition, the susceptibility rates of piperacillin/tazobactam and imipenem against E. coli were 86.3% and 100%. Meanwhile, the susceptibility rates of Streptococcus pneumoniae isolates to penicillin G, erythromycin, chloramphenicol, ceftriaxone and tetracycline were 68.8%, 0.0%, 87.5%, 81.3% and 0.0%, respectively. Gentamycin, ofloxacin, linezolid and vancomycin were identified as the most effective antibiotics for Streptococcus pneumoniae, each with susceptibility rates of 100%. It was notable that other emerging pathogens, such as Listeria monocytogenes and group D streptococcus, cannot be underestimated in meningitis.

Antibacterianos/uso terapêutico , Meningites Bacterianas/tratamento farmacológico , Doença Aguda , Bactérias/classificação , Bactérias/efeitos dos fármacos , Bactérias/isolamento & purificação , Líquido Cefalorraquidiano/microbiologia , Criança , Pré-Escolar , China , Farmacorresistência Bacteriana , Feminino , Humanos , Lactente , Recém-Nascido , Masculino , Meningites Bacterianas/líquido cefalorraquidiano , Testes de Sensibilidade Microbiana
World J Gastroenterol ; 20(18): 5511-8, 2014 May 14.
Artigo em Inglês | MEDLINE | ID: mdl-24833881


AIM: To identify potential biomarkers of primary gallbladder cancer (PGC). METHODS: Fresh PGC, cholecystitis and normal gallbladder tissue specimens collected from 10 patients, respectively, were subjected to comparative proteomic analysis. The proteomic patterns of PGC were compared with those of cholecystitis and normal gallbladder tissues using two-dimensional gel electrophoresis (2-DE). The differentially expressed proteins were then identified using a MALDI-TOF mass spectrometer (MS) and database searches. To further validate these proteins, 20 samples of PGC tissues and normal tumor-adjacent tissues were collected for Western blot, quantitative real-time PCR, and immunohistochemical staining assay. RESULTS: Seven differentially expressed protein spots were detected by 2-ED analysis by comparing the average maps of PGC, cholecystitis and normal gallbladder tissues. Six of the seven differentially expressed proteins were identified using MALDI-TOF MS, with three overexpressed and three underexpressed in PGC tissue. Protein levels of annexin A4 (ANXA4) were significantly elevated, and heat shock protein 90-beta (Hsp90ß) and dynein cytoplasmic 1 heavy chain 1 (Dync1h1) were decreased in PGC tissues relative to the normal tumor-adjacent tissues as shown by Western blot analysis. However, levels of actin, aortic smooth muscle and gamma-actin were unchanged. In addition, the mRNA levels of all 5 proteins showed similar changes to those of the protein levels (P < 0.01). Further validation by immunohistochemical analysis showed the upregulated expression of ANXA4 and decreased expression of Hsp90ß and Dync1h1 in the cytoplasm of PGC tissues relative to the normal tumor-adjacent tissues. CONCLUSION: Three proteins are identified as potential biomarkers of PGC using proteomic analysis. The functions of these proteins in the carcinogenesis of PGC remain to be studied.

Biomarcadores Tumorais/análise , Neoplasias da Vesícula Biliar/química , Proteínas de Neoplasias/análise , Proteômica , Idoso , Biomarcadores Tumorais/genética , Western Blotting , Estudos de Casos e Controles , Bases de Dados de Proteínas , Eletroforese em Gel Bidimensional , Feminino , Neoplasias da Vesícula Biliar/genética , Neoplasias da Vesícula Biliar/patologia , Regulação Neoplásica da Expressão Gênica , Humanos , Imuno-Histoquímica , Masculino , Pessoa de Meia-Idade , Proteínas de Neoplasias/genética , Valor Preditivo dos Testes , Proteômica/métodos , RNA Mensageiro/análise , Reação em Cadeia da Polimerase em Tempo Real , Reprodutibilidade dos Testes , Espectrometria de Massas por Ionização e Dessorção a Laser Assistida por Matriz
Zhongguo Dang Dai Er Ke Za Zhi ; 15(4): 264-7, 2013 Apr.
Artigo em Chinês | MEDLINE | ID: mdl-23607947


OBJECTIVE: To determine the frequency distribution and antibiotic resistance of pathogens isolated from the cerebrospinal fluid samples of children with bacterial meningitis (BM) and to provide a basis for the timely and effective treatment of childhood BM. METHODS: Retrospective analysis was performed on pathogens isolated from 5097 cerebrospinal fluid samples collected from children in Kunming Children's Hospital between January 2008 and June 2012, as well as drug sensitivity test results. Kirby-Bauer antibiotic testing was used to analyze the sensitivity of these pathogens to commonly used antibiotics. RESULTS: A total of 116 pathogen strains were detected from the 5097 cerebrospinal fluid samples, including 77 (66.4%) Gram-positive strains, 30 (25.9%) Gram-negative strains, and 9 (7.8%) fungal strains, with a positive rate of 2.28%. The six most frequently isolated pathogens were Staphylococcus epidermidis (32 strains, 27.6%), Streptococcus pneumoniae (15 strains, 12.9%), Escherichia coli (15 strains, 12.9%), Staphylococcus haemolyticus (9 strains, 7.8%), Cryptococcus neoformans (8 strains, 6.9%) and Staphylococcus aureus (6 strains, 5.2%). Coagulase-negative staphylococci was the predominant pathogen in neonates and young infants with BM, and its sensitivity rates to penicillin, erythromycin and clindamycin were lower than 40%. Streptococcus pneumoniae had a penicillin sensitivity rate of 13.4%, while sensitivity rates to erythromycin and clindamycin reached 60.0%. No Staphylococcus and Streptococcus pneumoniae pathogens resistant to vancomycin were found. Gram-negative bacilli had relatively high sensitivity rates to imipenem, meropenem, cefoperazone/sulbactam and cefepime. CONCLUSIONS: Gram-positive cocci are the predominant pathogens for childhood BM over the past five years. The detected pathogens develop high resistance to commonly used antibiotics. To prevent misdiagnosis, careful attention should be paid to BM caused by Cryptococcus neoformans.

Meningites Bacterianas/tratamento farmacológico , Criança , Pré-Escolar , Farmacorresistência Bacteriana , Feminino , Bactérias Gram-Negativas/efeitos dos fármacos , Bactérias Gram-Negativas/isolamento & purificação , Cocos Gram-Positivos/efeitos dos fármacos , Cocos Gram-Positivos/isolamento & purificação , Humanos , Lactente , Recém-Nascido , Masculino , Meningites Bacterianas/líquido cefalorraquidiano , Meningites Bacterianas/microbiologia , Estudos Retrospectivos