• Congenital anomalies and proximity to landfill sites.

      Boyle, E; Johnson, H; Kelly, A; McDonnell, R; Health Information Unit, Department of Public Health, Eastern Regional Health Authority, Dublin. (2004-01)
      The occurrence of congenital anomalies in proximity to municipal landfill sites in the Eastern Region (counties Dublin, Kildare, Wicklow) was examined by small area (district electoral division), distance and clustering tendancies in relation to 83 landfills, five of which were major sites. The study included 2136 cases of congenital anomaly, 37,487 births and 1423 controls between 1986 and 1990. For the more populous areas of the region 50% of the population lived within 2-3 km of a landfill and within 4-5 km for more rural areas. In the area-level analysis, the standardised prevalence ratios, empirical and full Bayesian modelling, and Kulldorff's spatial scan statistic found no association between the residential area of cases and location of landfills. In the case control analysis, the mean distance of cases and controls from the nearest landfill was similar. The odds ratios of cases compared to controls for increasing distances from all landfills and major landfills showed no significant difference from the baseline value of 1. The kernel and K methods showed no tendency of cases to cluster in relationship to landfills. In conclusion, congenital anomalies were not found to occur more commonly in proximity to municipal landfills.
    • Epitope discovery with phylogenetic hidden Markov models.

      Lacerda, Miguel; Scheffler, Konrad; Seoighe, Cathal; School of Mathematics, Statistics and Applied Mathematics, National University of Ireland, Galway, Ireland. (2010-05)
      Existing methods for the prediction of immunologically active T-cell epitopes are based on the amino acid sequence or structure of pathogen proteins. Additional information regarding the locations of epitopes may be acquired by considering the evolution of viruses in hosts with different immune backgrounds. In particular, immune-dependent evolutionary patterns at sites within or near T-cell epitopes can be used to enhance epitope identification. We have developed a mutation-selection model of T-cell epitope evolution that allows the human leukocyte antigen (HLA) genotype of the host to influence the evolutionary process. This is one of the first examples of the incorporation of environmental parameters into a phylogenetic model and has many other potential applications where the selection pressures exerted on an organism can be related directly to environmental factors. We combine this novel evolutionary model with a hidden Markov model to identify contiguous amino acid positions that appear to evolve under immune pressure in the presence of specific host immune alleles and that therefore represent potential epitopes. This phylogenetic hidden Markov model provides a rigorous probabilistic framework that can be combined with sequence or structural information to improve epitope prediction. As a demonstration, we apply the model to a data set of HIV-1 protein-coding sequences and host HLA genotypes.
    • Experimental validation of a Bayesian model of visual acuity.

      Dalimier, Eugénie; Pailos, Eliseo; Rivera, Ricardo; Navarro, Rafael; Applied Optics Group, School of Physics, National University of Ireland, Galway, Ireland. eugenie.dalimier@nuigalway.ie (2009)
      Based on standard procedures used in optometry clinics, we compare measurements of visual acuity for 10 subjects (11 eyes tested) in the presence of natural ocular aberrations and different degrees of induced defocus, with the predictions given by a Bayesian model customized with aberrometric data of the eye. The absolute predictions of the model, without any adjustment, show good agreement with the experimental data, in terms of correlation and absolute error. The efficiency of the model is discussed in comparison with image quality metrics and other customized visual process models. An analysis of the importance and customization of each stage of the model is also given; it stresses the potential high predictive power from precise modeling of ocular and neural transfer functions.