Search

Feature selection is one of the trending challenges in multi-label classification. In recent years a lot of methods have been proposed. However the existing approaches assume that all the features have the same cost. This assumption may be inappropriate when the acquisition of the feature values is costly. For example in medical diagnosis each diagnostic value extracted by a clinical test is associated with its own cost. In such cases it may be better to choose a model with an acceptable classification performance but a much lower cost. We propose a novel method which incorporates the feature cost information into the learning process. The method, named Cost-Sensitive Classifier Chains, combines classifier chains and penalized logistic regression with a modified elastic-net penalty which takes into account costs of the features. We prove the stability and provide a bound on generalization error of our algorithm. We also propose the adaptive version in which penalty factors are changing during fitting the consecutive models in the chain. The methods are applied on real datasets: MIMIC-II and Hepatitis for which the cost information is provided by experts. Moreover, we propose an experimental framework in which the features are observed with measurement errors and the costs depend on the quality of the features. The framework allows to compare the cost-sensitive methods on benchmark datasets for which the cost information is not provided. The proposed method can be recommended in a situation when one wants to balance low costs and high prediction performance.

Introduction: In vitro experiments have demonstrated that preconditioning primary neuronal cultures by temporary application of NMDA receptor antagonists induces long-term tolerance against lethal insults. In the present study we tested whether similar effects also occur in brain submitted to ischemia in vivo and whether the potential benefit outweighs the danger of enhancing the constitutive apoptosis in the developing brain. Material and methods: Memantine in pharmacologically relevant doses of 5 mg/kg or (+)MK-801 (3 mg/kg) was administered i.p. 24, 48, 72 and 96 h before 3-min global forebrain ischemia in adult Mongolian gerbils or prior to hypoxia/ ischemia in 7-day-old rats. Neuronal loss in the hippocampal CA1 in gerbils or weight deficit of the ischemic hemispheres in the rat pups was evaluated after 14 days. Also, the number of apoptotic neurons in the immature rat brain was evaluated. Results: In gerbils only the application of (+)MK-801 24 h before ischemia resulted in significant prevention of the loss of pyramidal neurons. In rat pups administration of (+)MK-801 at all studied times before hypoxia-ischemia, or pretreatment with memantine or with hypoxia taken as a positive control 48 to 92 h before the insult, significantly reduced brain damage. Both NMDA receptor antagonists equally reduced the number of apoptotic neurons after hypoxia-ischemia, while (+)MK-801-evoked potentiation of constitutive apoptosis greatly exceeded the effect of memantine. Conclusions: We ascribe neuroprotection induced in the immature rats by the pretreatment with both NMDA receptor antagonists 48 to 92 h before hypoxia-ischemia to tolerance evoked by preconditioning, while the neuroprotective effect of (+)MK-801 applied 24 h before the insults may be attributed to direct consequences of the inhibition of NMDA receptors. This is the first report demonstrating the phenomenon of inducing tolerance against hypoxia-ischemia in vivo in developing rat brain by preconditioning with NMDA receptor antagonists.

Perinatal brain insult mostly resulting from hypoxia-ischemia (H-I) often brings lifelong permanent disability, which has a major impact on the life of individuals and their families. The lack of progress in clinically-applicable neuroprotective strategies for birth asphyxia has led to an increasing interest in alternative methods of therapy, including induction of brain tolerance by pre- and particularly postconditioning. Hypoxic postconditioning represents a promising strategy for preventing ischemic brain damage. The aim of this study was to investigate the potential neuroprotective effect of hypobaric hypoxia (HH) postconditioning applied to 7-day old rats after H-I insult. The mild hypobaric conditions (0.47 atm) used in this study imitate an altitude of 5,000 m. We show that application of mild hypobaric hypoxia at relatively short time intervals (1-6 h) after H-I, repeated for two following days leads to significant neuroprotection, manifested by a reduction in weight loss of the ipsilateral hemisphere observed 14 days after H-I. HH postconditioning results in decrease in reactive oxygen species level observed in all experimental groups. The increase in superoxide dismutase activity observed after H-I is additionally enhanced by HH postconditioning applied 1 h after H-I. The increase observed 3 and 6 h after H-I was not statistically significant. Postconditioning with HH suppresses the glutathione concentration decrease evoked by H-I and increased glutathione peroxidase activity and this effect is not dependent on the time of postconditioning initiation. HH postconditioning had no effect on catalase activity. We show for the first time that HH postconditioning reduces brain damage resulting from H-I in immature rats and that the mechanism potentially involved in this effect is related to antioxidant defense mechanisms of immature brain.

The aim of this study was to establish the antagonistic effects of 1-methyl-1,2,3,4-tetrahydroisoquinoline (1MeTIQ) on NMDA receptors and its neuroprotective abilities on primary cultures of rat cerebellar granule cells exposed for 30 min to 250 or 100 μM glutamate. Neuronal viability was tested after 24 h with propidium iodide or calcein/ethidium homodimer-1 staining. The neuroprotective potential of 100, 250 or 500 μM 1MeTIQ was compared with established uncompetitive NMDA receptor antagonists, 0.5 Μm MK-801, or 5 μMmemantine. These substances were applied for 30 min either together with glutamate, 24 or 48 h before glutamate, or 0.5 h, 1 h and 3 h after exposure to the excitotoxin. The results demonstrated that MK-801, memantine and 500 μM 1MeTIQ induced an almost complete neuroprotection when co-applied with glutamate, but lower concentrations of 1MeTIQ were slightly less effective. Similar effects for 1MeTIQ and the established NMDA receptor antagonists were observed in the pretreatment experiments, even with a 48-h lag between the application of the tested substances and the excitotoxic challenge. In the post-treatment experiments, MK-801 and memantine and 500 μM 1MeTIQ applied up to 3 h after the exposure to glutamate significantly reduced the excitotoxic lesion, but 1MeTIQ in lower concentrations was ineffective. These results indicate that 1MeTIQ shares neuroprotective abilities with established uncompetitive NMDA receptor antagonists, which suggests that its inhibitory effect on NMDA receptors plays a key role in its anti-excitotoxic activity. Moreover, our data disclose a new mechanism of 1MeTIQ-evoked neuroprotection based on the induction of neuronal tolerance to excitotoxicity.

In the present study, we examined the effects of negative and positive allosteric modulators of metabotropic glutamate receptor 5 (mGluR5), fenobam and ADX47273, respectively, on brain damage induced by hypoxia-ischemia (H-I) in 7-day-old rats. The test drugs were administered intraperitoneally 10 min after H-I. Rectal body temperature was measured for 2.5 h after the insult. The number of apoptotic neurons in the immature rat brain was evaluated after 24 h. The wet weight of both hemispheres was determined 14 days after H-I, and its loss was used as an indicator of brain damage. In the vehicle-treated groups, H-I reduced the weight of the ipsilateral (ischemic) hemisphere by approximately 33% and sixfold increased the number of apoptotic cells in the cortex. Fenobam (10 mg/kg) and ADX47273 (5, 10, and 30 mg/kg) had no significant effect on brain damage, although application of fenobam at this dose significantly reduced the number of apoptotic cells. In contrast, fenobam (20 mg/kg) potentiated ischemic brain damage to 57.4% and had no effect on H-I-induced apoptosis. In all of the experimental groups, we detected no significant changes in the weight of the contralateral (control) hemisphere or the rectal temperature. In conclusion, in 7-day-old rats, the bidirectional modulation of mGluR5 by fenobam (10 mg/kg) and ADX47273 (all doses tested) did not result in significant changes in H-I-evoked brain damage, supporting our previous data indicating that also the antagonists of mGluR5 MPEP and MTEP, which reduce neuronal lesions in adult animals submitted to brain ischemia, were ineffective in 7-day-old rat pups.

Four virulent phages--PhiDl, PhiTl, PhiCYT21, and PhiOS6, infective on Sarothamnus scoparius rhizobia--were isolated from the soil and characterized for morphology, host range, rate of adsorption to bacterial cells, and genome size. New phages were separated into two morphological families: Siphoviridae with long, noncontractile tails (PhiDl, PhiTl) and Myoviridae with long, contractile tails (PhiCYT21, PhiOS6). They were also classified into two groups by a host specificity. One of them included viruses (PhiDl and PhiTl) that lysed S. scoparius bradyrhizobia and Bradyrhizobium sp. (Lupinus) strain Dl, and the second one comprised phages (PhiCYT21 and PhiOS6) that parasitized only Scotch broom native microsymbionts. Phages specific for S. scoparius rhizobia were differentiated not only by morphology and host range but also by a genome size that was in the range from 47,583 to 60,098 b.p.

Preconditioning is an experimental strategy for reducing ischemic brain damage. There are reports that brief exposure of neurons to NMDA-receptor antagonists may be an adequate preconditioning stressor. We studied effects of preconditioning of the cerebellar granule cells (CGC) in primary culture by 30-minute exposure to NMDA receptor antagonists 0.5 μM MK-801 or 5 μM memantine. CGC were challenged with oxygen and glucose deprivation (OGD) or excitotoxic glutamate and cell viability was tested 24 h later using calcein/ethidium homodimer-1 staining. We studied glutamate-induced increases in 45Ca uptake and in the intracellular Ca2+ level assessed with the fluorescent probe fluo-3. The number of living cells in OGD-treated cultures decreased by 42%. Preconditioning with MK-801 or memantine 24 h earlier reduced cell death to 8% and 30% and 48 h earlier to 27% and 33%, respectively. Pretreatment with MK-801 followed by the standard MK-801 wash out was slightly cytoprotective in a glutamate excitotoxicity test performed immediately; the protection increased significantly 24 h after preconditioning. In both cases the extensive wash out of MK-801 after preconditioning resulted in loss of cytoprotection. The increase in the intracellular Ca2+ level evoked by glutamate was decreased 24 h after preconditioning and even halved in the neuronal cultures 48 h after preconditioning with MK-801 and memantine. Glutamate-induced 45Ca uptake in these cells was decreased by 18%, irrespective of the time laps after preconditioning. These results demonstrate that preconditioning of CGC with NMDA receptor antagonists induces prolonged tolerance to OGD, which is accompanied by the reduction of glutamate-evoked calcium fluxes. The causal relationship between these effects may be suggested.

In vitro experiments have demonstrated that preconditioning primary neuronal cultures by temporary application of NMDA receptor antagonists induces long-term tolerance against lethal insults. In the present study we tested whether similar effects also occur in brain submitted to ischemia in vivo and whether the potential benefit outweighs the danger of enhancing the constitutive apoptosis in the developing brain.Memantine in pharmacologically relevant doses of 5 mg/kg or (+)MK-801 (3 mg/kg) was administered i.p. 24, 48, 72 and 96 h before 3-min global forebrain ischemia in adult Mongolian gerbils or prior to hypoxia/ischemia in 7-day-old rats. Neuronal loss in the hippocampal CA1 in gerbils or weight deficit of the ischemic hemispheres in the rat pups was evaluated after 14 days. Also, the number of apoptotic neurons in the immature rat brain was evaluated.In gerbils only the application of (+)MK-801 24 h before ischemia resulted in significant prevention of the loss of pyramidal neurons. In rat pups administration of (+)MK-801 at all studied times before hypoxia-ischemia, or pretreatment with memantine or with hypoxia taken as a positive control 48 to 92 h before the insult, significantly reduced brain damage. Both NMDA receptor antagonists equally reduced the number of apoptotic neurons after hypoxia-ischemia, while (+)MK-801-evoked potentiation of constitutive apoptosis greatly exceeded the effect of memantine.We ascribe neuroprotection induced in the immature rats by the pretreatment with both NMDA receptor antagonists 48 to 92 h before hypoxia-ischemia to tolerance evoked by preconditioning, while the neuroprotective effect of (+)MK-801 applied 24 h before the insults may be attributed to direct consequences of the inhibition of NMDA receptors. This is the first report demonstrating the phenomenon of inducing tolerance against hypoxia-ischemia in vivo in developing rat brain by preconditioning with NMDA receptor antagonists.

Colorectal carcinoma is the most common gastrointestinal malignancy in the world. The survival of patients with colorectal cancer has not varied appreciably in recent years. The knowledge that genetic factors could play a role in the etiology and prognosis of patients with sporadic colorectal cancer has opened up new lines of research. The aim of the study was to evaluate a group of patients with colorectal cancer and the possible influence on the prognosis of immunohistochemical hMLH1, hMSH2 and CD34 expression.37 cases of colorectal cancer were included in this study. HMLH1, hMSH2 and CD34 protein expression in tumor tissue sections was determined using immunohistochemical staining with specific monoclonal antibodies. A total of 37 tumors were analyzed for histopathology. The immunoexpression status of MMR genes was correlated with cancer stage, location, histology, gender and CD34 immunoexpression.Of the 37 cases immunohistochemical analysis showed 13 (35.1%) had loss of expression of hMLH1, 5 (13.5%) had loss of expression of hMSH2, and five cases had loss of expression of both proteins. Multivariate analysis showed that right side dominance (p0.05) poorly differentiated and male (p0.05) were associated with loss of expression of hMLH1 or hMSH2. Loss of expression of hMLH1 or hMSH2 correlated with low CD34 expression (p0.05). CD34 immunoexpression was higher in the left side tumor location, stage D Dukes'a (p0.05) and G2 (p0.05).These results confirm that sporadic colorectal MMR-defective adenocarcinomas display certain specific morphological characteristics. However, these pathological features are not sufficiently predictive and immunohistochemistry is needed to identify such tumours accurately. The researches on angiogenesis of cancer cells can be used as prognostic factors for the patients with colorectal cancer.

The role of perinuclear antineutrophil cytoplasmic antibodies (pANCA) and anti-Saccharomyces cerevisiae antibodies (ASCA) assessment in inflammatory bowel disease (IBD) diagnosis and differentiating is still imprecise and controversial. The aim of the study was to determine the accuracy of pANCA and ASCA in patients with IBD subgroups.The study was performed in 125 patients: 71 patients with ulcerative colitis (UC), 31 with Crohn's disease (CD) and 23 with indeterminate Colitis (IC). Control group consists of 45 patients with functional intestinal disorders. pANCA and ASCA (IgA and IgG) were measured with ELISA, using commercial antibody panel.In UC patients the prevalence of pANCA was 68%, which was significantly higher than in CD-29%. ASCA were found significantly more often in CD-80.6% than in UC patients-26.8%. Sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) of pANCA for UC diagnosis was 68%, 84%, 75% and 78%; and ASCA for CD: 81%, 78%, 45,5% and 95%, respectively. The combined use of these two markers gave changes in diagnostic accuracy: pANCA+/ASCA- in UC: 42%, 100%, 100% and 43%, and for pANCA-/ASCA+ in CD: 52%, 98.6% 94% and 82%, respectively.The specificity of these combined markers tends to be higher than sensitivity, what made them more useful in the differentiation of the IBD subtypes rather than population screening. The characteristic IC serotype pANCA(-)ASCA(-) leads to further controversies about origin of this IBD subtype.

The aim of this study was to compare the results of two methods: disc-diffusion test and ATB STAPH 5 (version 2000) test applied for determination of methicillin-resistant staphylococci. One hundred and twenty S. aureus strains isolated from clinical specimens in microbiological laboratory of Medical University in Warsaw were investigated. Identification and drug resistance were determined using automatic ATB Expression System. We compared the results of methicillin-resistant S. aureus detection by disc-diffusion method (oxacillin 1 microg) and ATB STAPH 5; version 2000 (bioMérieux sa). In the case of 116 strains S. aureus identical results were obtained in both methods, 4 strains gave inconsistent results in these two methods. Results that were obtained show that the ATB STAPH 5 (version 2000) test is as useful for routine determination of MRSA and MSSA strains as the disc-diffusion method, and may be used alternatively.