Browsing by Author "Lacerda, Miguel"
Now showing 1 - 9 of 9
Results Per Page
Sort Options
- ItemOpen Access9β Polymorphism of the Glucocorticoid Receptor Gene Appears to Have Limited Impact in Patients with Addison’s Disease(Public Library of Science, 2014) Ross, Ian Louis; Dandara, Collet; Swart, Marelize; Lacerda, Miguel; Schatz, Desmond; Blom, Dirk JacobusBACKGROUND: Addison’s disease (AD) has been associated with an increased risk of cardiovascular disease. Glucocorticoid receptor polymorphisms that alter glucocorticoid sensitivity may influence metabolic and cardiovascular risk factors in patients with AD. The 9β polymorphism of the glucocorticoid receptor gene is associated with relative glucocorticoid resistance and has been reported to increase the risk of myocardial infarction in the elderly. We explored the impact of this polymorphism in patients with AD. Materials and METHODS: 147 patients with AD and 147 age, gender and ethnicity matched healthy controls were recruited. Blood was taken in a non-fasted state for plasma lipid determination, measurement of cardiovascular risk factors and DNA extraction. RESULTS: Genotype data for the 9β polymorphism was available for 139 patients and 146 controls. AD patients had a more atherogenic lipid profile characterized by an increase in the prevalence of small dense LDL (p = 0.003), increased triglycerides (p = 0.002), reduced HDLC (p<0.001) an elevated highly sensitive C-reactive protein (p = 0.01), compared with controls. The 9β polymorphism (at least one G allele) was found in 28% of patients and controls respectively. After adjusting for age, gender, ethnicity, BMI and hydrocortisone dose per metre square of body surface area in patients, there were no significant metabolic associations with this polymorphism and hydrocortisone doses were not higher in patients with the polymorphism. CONCLUSIONS: This study did not identify any associations between the 9β polymorphism and cardiovascular risk factors or hydrocortisone dose and determination of this polymorphism is therefore unlikely to be of clinical benefit in the management of patients with AD.
- ItemOpen AccessAnti-retroviral therapy increases the prevalence of dyslipidemia in South African HIV-infected patients(Public Library of Science, 2016) Dave, Joel A; Levitt, Naomi S; Ross, Ian L; Lacerda, Miguel; Maartens, Gary; Blom, DirkPurpose Data on the prevalence of dyslipidaemia and associated risk factors in HIV-infected patients from sub-Saharan Africa is sparse. We performed a cross-sectional analysis in a cohort of HIV-infected South African adults. METHODS: We studied HIV-infected patients who were either antiretroviral therapy (ART)-naive or receiving non-nucleoside reverse transcriptase inhibitor (NNRTI)-based or protease inhibitor (PI)-based ART. Evaluation included fasting lipograms, oral glucose tolerance tests and clinical anthropometry. Dyslipidemia was defined using the NCEP ATPIII guidelines. RESULTS: The median age of the participants was 34 years (range 19-68 years) and 78% were women. The prevalence of dyslipidemia in 406 ART-naive and 551 participants on ART was 90.0% and 85%, respectively. Low HDL-cholesterol (HDLC) was the most common abnormality [290/406 (71%) ART-naïve and 237/551 (43%) ART- participants]. Participants on ART had higher triglycerides (TG), total cholesterol (TC), LDL-cholesterol (LDLC) and HDLC than the ART-naïve group. Severe dyslipidaemia, (LDLC> 4.9 mmol/L or TG >5.0 mmol/L) was present in <5% of participants. In multivariate analyses there were complex associations between age, gender, type and duration of ART and body composition and LDLC, HDLC and TG, which differed between ART-naïve and ART-participants. CONCLUSION: Participants on ART had higher TG, TC, LDLC and HDLC than those who were ART-naïve but severe lipid abnormalities requiring evaluation and treatment were uncommon.
- ItemOpen AccessCollaborative Genre Tagging(2020) Leslie, James; Lacerda, MiguelRecommender systems (RS) are used extensively in online retail and on media streaming platforms to help users filter the plethora of options at their disposal. Their goal is to provide users with suggestions of products or artworks that they might like. Content-based RS's make use of user and/or item metadata to predict user preferences, while collaborative-filtering (CF) has proven to be an effective approach in tasks such as predicting movie or music preferences of users in the absence of any metadata. Latent factor models have been used to achieve state-of-the-art accuracy in many CF settings, playing an especially large role in beating the benchmark set in the Netflix Prize in 2008. These models learn latent features for users and items to predict the preferences of users. The first latent factor models made use of matrix factorisation to learn latent factors, but more recent approaches have made use of neural architectures with embedding layers. This master's dissertation outlines collaborative genre tagging (CGT), a transfer learning application of CF that makes use of latent factors to predict genres of movies, using only explicit user ratings as model inputs.
- ItemOpen AccessFine-scale drivers of African Penguin prey dynamics in Algoa Bay, South Africa, and their impacts on penguin foraging ecology(2016) Mcinnes, Alistair McIntyre; Ryan, Peter; Pichegru, Lorien; Lacerda, MiguelAfrican Penguins (Spheniscus demersus) have undergone a dramatic decrease in their population since the turn of this century prompting the up-grading of their conservation status to 'endangered'. There is growing evidence that variation in the availability of their principle prey, pelagic shoaling fish, are driving this trend. This prey variability is driven by oceanographic factors as well as commercial purse-seine fishing operations. To isolate the direct impacts of fishing on the foraging performance of African Penguins, the primary oceanographic drivers of fish distribution and abundance were investigated by conducting fine-scale pelagic fish surveys around two of the largest breeding colonies of African Penguins in Algoa Bay, St Croix and Bird islands, between 2011 and 2014. Quantification of fish parameters were facilitated by a novel method using a recreational fishfinder and calibrating this instrument to a conventional scientific device. The specific types of fish assemblages selected for by African Penguins were then evaluated by looking at the correspondence in associations of fish and penguins recorded at sea using both counts and locations of foraging birds tracked simultaneously during a subset of fish surveys. Activity budgets of penguins calculated from these simultaneous deployments were modelled against the abundance of their prey to elucidate hypothesised functional relationships. Finally, the direct influence of purse-seine fishing on both targeted fish assemblages and penguin activity budgets were assessed by modelling interactions between known physical drivers of targeted fish assemblages and different levels of cumulative catches. Physical drivers of the three-dimensional distribution and abundance of fish varied between colonies with primary production playing the most important role around Bird Island but having little influence on fish around St Croix Island where factors associated with surface and sea-profile temperatures had a stronger influence. Results of both penguin count and track data highlight the importance of the vertical distribution of prey to the distribution of foraging African Penguins with the abundance of these assemblages having a significant influence on this species' activity budgets. Evidence for local depletion of pelagic fish was demonstrated for the waters around St Croix Island and the effects of purse-seine fishing on African Penguin foraging effort were significant when controlling for natural drivers of prey distribution. Results of this research should be applied to current conservation measures, most notably alleviating direct competition by purse-seine fishing operations during periods of reduced primary productivity and when the abundance of targeted fish aggregations are significantly diminished three months prior to and during the onset of the African Penguin breeding season.
- ItemOpen AccessIdentification of broadly neutralizing antibody epitopes in the HIV-1 envelope glycoprotein using evolutionary models(BioMed Central Ltd, 2013) Lacerda, Miguel; Moore, Penny; Ngandu, Nobubelo; Seaman, Michael; Gray, Elin; Murrell, Ben; Krishnamoorthy, Mohan; Nonyane, Molati; Madiga, Maphuti; Wibmer, Constantinos; Sheward, Daniel; Bailer, Robert; Gao, Hongmei; Greene, Kelli; Karim, Salim S; MBACKGROUND:Identification of the epitopes targeted by antibodies that can neutralize diverse HIV-1 strains can provide important clues for the design of a preventative vaccine. METHODS: We have developed a computational approach that can identify key amino acids within the HIV-1 envelope glycoprotein that influence sensitivity to broadly cross-neutralizing antibodies. Given a sequence alignment and neutralization titers for a panel of viruses, the method works by fitting a phylogenetic model that allows the amino acid frequencies at each site to depend on neutralization sensitivities. Sites at which viral evolution influences neutralization sensitivity were identified using Bayes factors (BFs) to compare the fit of this model to that of a null model in which sequences evolved independently of antibody sensitivity. Conformational epitopes were identified with a Metropolis algorithm that searched for a cluster of sites with large Bayes factors on the tertiary structure of the viral envelope. RESULTS: We applied our method to ID50 neutralization data generated from seven HIV-1 subtype C serum samples with neutralization breadth that had been tested against a multi-clade panel of 225 pseudoviruses for which envelope sequences were also available. For each sample, between two and four sites were identified that were strongly associated with neutralization sensitivity (2ln(BF)>6), a subset of which were experimentally confirmed using site-directed mutagenesis. CONCLUSIONS: Our results provide strong support for the use of evolutionary models applied to cross-sectional viral neutralization data to identify the epitopes of serum antibodies that confer neutralization breadth.
- ItemOpen AccessIdentifying predictors of evolutionary dispersion with phylogeographic generalised linear models(2017) Wolff-Piggott, Timothy; Lacerda, Miguel; Murrell, BenDiscrete phylogeographic models enable the inference of the geographic history of biological organisms along phylogenetic trees. Frequently applied in the context of epidemiological modelling, phylogeographic generalised linear models were developed to allow for the evaluation of multiple predictors of spatial diffusion. The standard phylogeographic generalised linear model formulation, however, assumes that rates of spatial diffusion are a noiseless deterministic function of the set of covariates, admitting no other unobserved sources of variation. Under a variety of simulation scenarios, we demonstrate that the lack of a term modelling stochastic noise results in high false positive rates for predictors of spatial diffusion. We further show that the false positive rate can be controlled by including a random effect term, thus allowing unobserved sources of rate variation. Finally, we apply this random effects model to three recently published datasets and contrast the results of analysing these datasets with those obtained using the standard model. Our study demonstrates the prevalence of false positive results for predictors under the standard phylogeographic model in multiple simulation scenarios and, using empirical data from the literature, highlights the importance of a model accounting for random variation.
- ItemOpen AccessRecreational fish-finders - an inexpensive alternative to scientific echo-sounders for unravelling the links between marine top predators and their prey(Public Library of Science, 2015) McInnes, Alistair M; Khoosal, Arjun; Murrell, Ben; Merkle, Dagmar; Lacerda, Miguel; Nyengera, Reason; Coetzee, Janet C; Edwards, Loyd C; Ryan, Peter G; Rademan, JohanStudies investigating how mobile marine predators respond to their prey are limited due to the challenging nature of the environment. While marine top predators are increasingly easy to study thanks to developments in bio-logging technology, typically there is scant information on the distribution and abundance of their prey, largely due to the specialised nature of acquiring this information. We explore the potential of using single-beam recreational fish-finders (RFF) to quantify relative forage fish abundance and draw inferences of the prey distribution at a fine spatial scale. We compared fish school characteristics as inferred from the RFF with that of a calibrated scientific split-beam echo-sounder (SES) by simultaneously operating both systems from the same vessel in Algoa Bay, South Africa. Customized open-source software was developed to extract fish school information from the echo returns of the RFF. For schools insonified by both systems, there was close correspondence between estimates of mean school depth (R 2 = 0.98) and school area (R 2 = 0.70). Estimates of relative school density (mean volume backscattering strength; S v ) measured by the RFF were negatively biased through saturation of this system given its smaller dynamic range. A correction factor applied to the RFF-derived density estimates improved the comparability between the two systems. Relative abundance estimates using all schools from both systems were congruent at scales from 0.5 km to 18 km with a strong positive linear trend in model fit estimates with increasing scale. Although absolute estimates of fish abundance cannot be derived from these systems, they are effective at describing prey school characteristics and have good potential for mapping forage fish distribution and relative abundance. Using such relatively inexpensive systems could greatly enhance our understanding of predator-prey interactions.
- ItemOpen AccessRecurrent neural network language models in the context of under-resourced South African languages(2018) Scarcella, Alessandro; Lacerda, MiguelOver the past five years neural network models have been successful across a range of computational linguistic tasks. However, these triumphs have been concentrated in languages with significant resources such as large datasets. Thus, many languages, which are commonly referred to as under-resourced languages, have received little attention and have yet to benefit from recent advances. This investigation aims to evaluate the implications of recent advances in neural network language modelling techniques for under-resourced South African languages. Rudimentary, single layered recurrent neural networks (RNN) were used to model four South African text corpora. The accuracy of these models were compared directly to legacy approaches. A suite of hybrid models was then tested. Across all four datasets, neural networks led to overall better performing language models either directly or as part of a hybrid model. A short examination of punctuation marks in text data revealed that performance metrics for language models are greatly overestimated when punctuation marks have not been excluded. The investigation concludes by appraising the sensitivity of RNN language models (RNNLMs) to the size of the datasets by artificially constraining the datasets and evaluating the accuracy of the models. It is recommended that future research endeavours within this domain are directed towards evaluating more sophisticated RNNLMs as well as measuring their impact on application focused tasks such as speech recognition and machine translation.
- ItemOpen AccessA topic model based approach to inferring episodic directional selection in protein coding sequences(2015) Sadiq, Hassan Taiwo; Lacerda, MiguelPathogens, such as HIV and influenza, evolve in response to the selective pressures of their host environments accumulating changes in their genomes that offer fitness benefits. This selective pressure is characterised by three properties: (1.) it is episodic, tracking changes in the adaptive immune response and drug therapy, (2.) it is directional in that only particular amino acid substitutions are favoured and (3.) it varies between genomic loci. Most previous models have ignored or inadequately addressed some of these phenomena. This work extends recent approaches to modelling episodic directional selection acting on protein-coding sequences. We use inference techniques within the topic model framework to identify loci evolving under natural selection. A notable example of such techniques are the variational Bayesian methods. We show that our approach performs well in terms of specificity and power, and demonstrate its utility by applying it to some real datasets of HIV sequences.