In practice, most approaches for text error detection and correction are based on a conventional domain-dependent background dictionary that represents a fixed and static collection of correct words of a given language and, as a result, satisfactory correction can only be achieved if the dictionary covers most tokens of the underlying correct text. Again, most approaches for text correction are for only one or at best a very few types of errors. The purpose of this thesis is to propose an unsupervised approach to detecting and correcting text errors, that can compete with supervised approaches and answer the following questions: Can an unsupervised approach efficiently detect and correct a text containing multiple errors of both syntactic and semantic nature? What is the magnitude of error coverage, in terms of the number of errors that can be corrected? We conclude that (1) it is possible that an unsupervised approach can efficiently detect and correct a text containing multiple errors of both syntactic and semantic nature. Error types include: real-word spelling errors, typographical errors, lexical choice errors, unwanted words, missing words, prepositional errors, article errors, punctuation errors, and many of the grammatical errors (e.g., errors in agreement and verb formation). (2) The magnitude of error coverage, in terms of the number of errors that can be corrected, is almost double of the number of correct words of the text. Although this is not the upper limit, this is what is practically feasible. We use engineering approaches to answer the first question and theoretical approaches to answer and support the second question. We show that finding inherent properties of a correct text using a corpus in the form of an n-gram data set is more appropriate and practical than using other approaches to detecting and correcting errors. Instead of using rule-based approaches and dictionaries, we argue that a corpus can effectively be used to infer the properties of these types of errors, and to detect and correct these errors. We test the robustness of the proposed approach separately for some individual error types, and then for all types of errors. The approach is language-independent, it can be applied to other languages, as long as n-grams are available. The results of this thesis thus suggest that unsupervised approaches, which are often dismissed in favor of supervised ones in the context of many Natural Language Processing (NLP) related tasks, may present an interesting array of NLP-related problem solving strengths.

<p>Risk factors such as intrauterine and vaginal infection put pregnant women at risk for delivering preterm. Bacterial vaginosis (BV) is a polymicrobial clinical syndrome commonly diagnosed in women of reproductive age, with women of African descent with low socioeconomic status and previous preterm delivery at high risk. Although frequently isolated from healthy women,&nbsp / Gardnerella vaginalis has been most frequently associated with BV. There is limited data available on the prevalence of BV in Southern Africa / therefore, we embarked on a study to determine the&nbsp / prevalence of BV and G. vaginalis in predominantly black communities in the Western Cape, in order to establish the role of G. vaginalis in BV. Women attending various Maternity and Obstetrics&nbsp / units (MOU) in the Cape Peninsula with and without a history of pre-term delivery (PTD) were invited to participate in the study. Several factors were statistically associated with pregnancy history,&nbsp / including location of study population, parity, smoking and presence of clinical symptoms. The presence of G. vaginalis was determined by culture in 51.7% of the preterm delivery group (PTDG)&nbsp / and 44% of the full-term delivery group (FTDG) women. BV was detected in 31.13% of PTDG and 23.67% of FTDG by Gram stained analysis according to Nugent scoring criteria, with age and HIV&nbsp / status posing as risk factors. When comparing PTDG and FTDG for an association between the presence of G. vaginalis and BV, a stronger association was observed in the PTDG but it was not statistically significant. In both PTDG and FTDG, G. vaginalis was isolated significantly more often in women diagnosed with BV at 24.5% (p &lt / 0.05). Antibiogram studies revealed both Metronidazole and Clindamycin resistant strains of G. vaginalis. G. vaginalis Biotype 7 is specifically associated with BV, while Biotype 2 appears to be associated with BV in women with a history&nbsp / of PTD. Accuracy of diagnostic tools were tested and it was determined that Nugent scoring is more sensitive in diagnosing BV (76.04%), but culture for G. vaginalis is more specific (83.21%). Although this study was limited in that we were unable to follow-up pregnancy outcomes, we were able to confirm the perceived role of G. vaginalis in BV.&nbsp / </p>

DE LA TESIS:Desde la aparición de las quinolonas, las resistencias bacterianas a las mismas han evolucionado de forma paralela. Conforme han ido apareciendo nuevas moléculas con mayor capacidad bactericida y mejores parámetros farmacodinámicos y farmacocinéticos, también se han ido identificando nuevos y más sofisticados mecanismos de resistencia, que se han ido adaptando a las necesidades de cada momento. Es por ello que se hace necesaria una nueva metodología para el diseño y la síntesis de nuevos agentes antibacterianos con el objetivo de aumentar su potencia antibacteriana y, al mismo tiempo, disminuir la probabilidad de generar a corto plazo nuevos mecanismos de resistencia. Con este fin, es fundamental conocer todos los mecanismos de resistencia y cuál es, en detalle, su mecanismo de acción, para poder diseñar nuevas moléculas capaces de, manteniendo su capacidad bactericida, eludir dichos mecanismos.En esta tesis nos propusimos como objetivos fundamentales, en primer lugar, investigar las bases moleculares de los mecanismos de resistencia a quinolonas en bacterias Gram-negativas. En concreto, en Escherichia coli, Yersinia enterocolitica y Citrobacter freundii. Por otro lado, se llevaron a cabo cálculos de acoplamiento o "docking" mediante la utilización de programas informáticos con el fin de incrementar nuestros conocimientos respecto al modelo de interacción entra ADN-ADN girasa-quinolona, para de esta manera mejorar nuestra comprensión de los mecanismos de resistencia a fluoroquinolonas asociados a las mutaciones más comúnmente encontradas en el gen gyrA. Finalmente, el último de los objetivos que nos propusimos fue el de diseñar, sintetizar y evaluar diferentes derivados de ciprofloxacino y norfloxacino con el fin de encontrar entre ellos alguno que tuviera la capacidad de interaccionar con la enzima ADN girasa, poseyendo ésta uno o dos cambios aminoacídicos en su subunidad A.

In Streptococcus pyogenes, the protein SiaA (HtsA) is part of a heme uptake pathway system and involved in heme transfer from Shp to the ABC transporter. SiaA mutants, in which alanine replaces the axial histidine (H229) and methionine (M79) ligands, as well as a lysine (K61) and cysteine (C58) located near the heme propionates, are reported. Studies on a mutant of a cysteine expected to be at a distance from the propionates (C47A) are also reported. The coordination state and spin state of the selected mutants were determined via Resonance Raman studies. The pKa values of mutants ranged from 9.0 to 9.4, which were close to the pKa of the WT SiaA (9.7). The midpoint reduction potential of lysine (K61A) mutant was determined by spectroelectrochemical titration to be 61 ± 3 mV vs. SHE, similar to the WT protein (68 ± 3 mV). The addition of guanidinium hydrochloride resulted in protein denaturation that could show more than one process and occurred over days. The ease of protein unfolding was directly related to the extent of interaction of the residues with the heme: changes in the axial ligands resulted in far greater changes in heme protein stability than changes in the residues near the heme propionates. The causative agent of diphtheriae, Corynebacterium diphtheriae, imports heme via an ABC uptake transporter. In this research, two of the five proteins in the heme uptake pathway of C. diphtheriae were studied. These proteins were HmuT, lipoprotein component of the ABC transporter, and HtaA, the heme receptor. UV-visible spectroscopy and fluorescence spectroscopy showed that HmuT protein as isolated bound a porphyrin, rather than heme. Electrospray ionization mass spectrometry (ESI-MS) studies suggested that two tetrapyrroles were bound. To assess stability of this protein towards heme release, thermal denaturation studies were performed. For HtaA, UV-visible and fluorescence spectroscopy also showed the protein as isolated was also bound a porphyrin, rather than heme. Homology studies showed that HtaA protein is quiet distant from homologous heme uptake proteins and could be a member of novel heme binding domain family.

Sortase A is a housekeeping enzyme of Gram-positive bacteria that catalyses the anchoring of surface proteins to the bacterial peptidoglycan. The enzyme works to establish an interaction between bacteria and host cells and is essential for pathogenesis. This makes Sortase A a potential suitable target for inhibition, in order to treat bacterial infections. In this degree project Sortase A from Staphylococcus aureus was explored and potential inhibitors were investigated by performing enzyme activity and bacterial binding assays. A robust FRET assay was developed and optimized for a recombinant version of the enzyme and serves as a good starting point for studying inhibition.

Early detection of infection, identification of microorganism, and correct choice of antibiotics are critical in the management of sepsis. Quantitative real-time polymerase chain reaction (RT-PCR) has the potential to improve the timeliness, sensitivity, and accuracy of detecting pathogens. In this study we utilize this method to detect the enteric bacterial counts in the blood from patients with systemic inflammatory response syndrome (SIRS) in the emergency department (ED). The universal primers utilized in RT-PCR are specific for 23S ribosomal DNA (rDNA) and wec F gene. The results show that in SIRS patients with positive culture results from specimen collected within 10 days after presenting to ED, and patients surviving for less than 28 days, the serum bacterial DNA load of enteric Gram negative bacilli is higher. In SIRS patients with shock, patients fulfilling both white blood cell counts and respiratory criteria of SIRS, and patients fulfilling both white blood cell counts and respiratory criteria of SIRS with Acute Physiology and Chronic Health Evaluation II score more than 20, the serum bacterial DNA load of enteric Gram negative bacilli and 28-day mortality are both higher. These results suggest that bacterial translocation may happen in patients with SIRS and may be related to higher mortality in patients with SIRS.

No description available.

n the problem of time delay estimation (TDE), the desired source signals of interest are correlated and with a specific spectral distribution. In such cases, the convergence speed using the conventional approaches, viz., time domain adaptive constrained and unconstrained LMS TDE algorithms, becomes slowly and the performance of TDE will be degraded, dramatically. In fact, the convergence rate depends highly on the distribution of spectral density of the desired signal sources. Also, the performance of TDE is affected by the background noises, accordingly. To circumvent the problem described above, in this thesis, a transformed domain adaptive constrained filtering scheme, refers to the constrained adaptive DCT-LMS algorithm, for TDE is devised. We show that this new proposed constrained algorithm, with the so-called direct delay estimation formula, for non-integer TDE does perform better than the conventional time domain adaptive constrained and unconstrained LMS TDE algorithms and the unconstrained adaptive DCT-LMS TDE algorithm. Finally, to further reduce the spread of eigenvalue in the unconstrained adaptive DCT-LMS algorithm, the Gram-Schmidt orthogonalizer approach realizing by the adaptive Escalator is investigated. It indicates that bias of TDE will occur without using the constraint of weight vector. That is, it could not be used to alleviate the effect due to background noises.

Thèse de doctorat : Aspects moléculaires et cellulaires de la biologie : Strasbourg 1 : 2007. / Thèse soutenue sur un ensemble de travaux. Titre provenant de l'écran-titre. Bibliogr. p. 152-162.

Hard-to-heal ulcers are resource intensive due to the fact that they are difficult to treat and especially vulnerable to bacterial invasion. The bacterial culture contaminating these wounds often consist of several different bacterial organisms that originate from endogenous sources. Necrotic material in ischemic ulcers provide nutrition which support bacterial reproduction, increasing the risk of infection. Determining causative pathogen in infected ulcers proves to be difficult when culturing swab samples, however Staphylococcus aureus and hemolytic streptococci generally act as primary pathogens.     The aim of the study was to investigate if the detection rate increased for S. aureus and hemolytic streptococci when culturing swab samples from ulcers on Columbia CNA; a media selective for gram-positive bacteria. In the experimental procedure the inhibitory action of CNA upon gram-negative bacterial growth was evaluated, using simulated ulcer samples (n=6) containing bacterial quality control strains in arbitrary concentrations. Additionally, patient samples (n=51) were cultured and screened for primary pathogens to investigate differences in the detection rate for CNA and the current culture media; Blood agar, Chocolate agar, Gentian violet blood agar and CLED agar.    Results from simulated ulcer samples showed excellent inhibitory function regarding the antibiotic substances of the CNA agar. Culturing patient samples from lower leg- and pressure ulcers on CNA, provided indications of diverse circumstances yielding higher respectively lower detection rate concerning S. aureus and hemolytic streptococci. Samples containing mixed flora with gram-negative bacteria generated higher detection rate and samples containing S. aureus yielded a lower detection rate when culturing on CNA, compared with that of the routine method.