Subspace-based techniques and applications

Este trabalho focou-se no estudo de técnicas de sub-espaço tendo em vista as aplicações seguintes: eliminação de ruído em séries temporais e extracção de características para problemas de classificação supervisionada. Foram estudadas as vertentes lineares e não-lineares das referidas técnicas tendo como ponto de partida os algoritmos SSA e KPCA. No trabalho apresentam-se propostas para optimizar os algoritmos, bem como uma descrição dos mesmos numa abordagem diferente daquela que é feita na literatura. Em qualquer das vertentes, linear ou não-linear, os métodos são apresentados utilizando uma formulação algébrica consistente. O modelo de subespaço é obtido calculando a decomposição em valores e vectores próprios das matrizes de kernel ou de correlação/covariância calculadas com um conjunto de dados multidimensional. A complexidade das técnicas não lineares de subespaço é discutida, nomeadamente, o problema da pre-imagem e a decomposição em valores e vectores próprios de matrizes de dimensão elevada. Diferentes algoritmos de préimagem são apresentados bem como propostas alternativas para a sua optimização. A decomposição em vectores próprios da matriz de kernel baseada em aproximações low-rank da matriz conduz a um algoritmo mais eficiente- o Greedy KPCA. Os algoritmos são aplicados a sinais artificiais de modo a estudar a influência dos vários parâmetros na sua performance. Para além disso, a exploração destas técnicas é extendida à eliminação de artefactos em séries temporais biomédicas univariáveis, nomeadamente, sinais EEG.

The impact of communication disability on quality of life of people with aphasia and their caregivers

Aphasia is a chronic acquired communication disorder that may change people with aphasia (PWA) and their caregivers’ lives for ever. Social and emotional changes are frequently reported by both, although the impact of these changes in quality of life (QOL) needs further research. This study identifies predictors and variables that influence PWA’s and their caregivers’ QOL and social relationships (SR). A cross-sectional descriptive, correlational and comparative study was undertaken with 255 individuals from Portuguese general population (mean age 43 years, range 25-84 years; 148 females, 107 males), 25 PWA (mean age 54 years, range 20-71; 12 females and 13 males), and 25 caregivers (mean age 51 years, range 26-73; 17 females and 8 males). All the participants completed the World Health Quality of Life Bref instrument, the SR domain of the World Health Organization Quality of Life – 100 scale, and the Center for Epidemiologic Studies Depression Scale. Participants with aphasia completed the Lisbon Aphasia Assessment Battery, Barthel Index, Frenchay Activities Index, Communication Disability Profile and the Modified Mini-Mental State which evaluated language disability, aphasia coefficient, activities, participation and cognition. In addition, caregivers completed the Carers’ Assessment of Managing Index to assess coping strategies. PWA and their caregivers are less satisfied with their QOL and SR than Portuguese general population; whereas PWA have the worse QOL and SR. Thus, aphasia seems to impact negatively in their QOL and SR. Emotional status has great importance for QOL and SR among the three groups. PWA’s activities and participation have great impact in both PWA’s and caregivers’ QOL and SR. Additionally, emotional status and participation are the best predictors of PWA’s QOL. Along with these two variables, activities of the PWA are the best predictors of caregivers’ QOL. Participation is the best predictor of PWA’s SR; emotional status and number of cohabitants are best predictors for caregivers’ SR. Aphasia assessment and intervention should take into account all the factors that influence PWA’s and caregivers’ QOL and SR so the central goal of enhancing it can be achieved. These results are important for identifying and planning support needs and are useful in the orientation of the activities carried out by the service providers allowing the adjustment of health programs and policies based on people’s real life needs.

Macroinvertebrate communities and habitat structure in least-impacted hill streams of central Portugal

We surveyed macroinvertebrate communities in 31 hill streams in the Vouga River and Mondego River catchments in central Portugal. Despite applying a "least-impacted" criterion, channel and bank management was common, with 38% of streams demonstrating channel modification (damming) and 80% with evidence of bank modification. Principal component analysis (PCA) at the family and species level related the macroinvertebrates to habitat variables derived at three spatial scales -- site (20 m), reach (200 m), and catchment. Variation in community structure between sites was similar at the species and family level and was statistically related to pH, conductivity, temperature, flow, shade, and substrate size at the site scale; channel and bank habitat and riparian vegetation and land-use at the reach scale; and altitude and slope at the catchment scale. While the effects of river management were apparent in various ecologically important habitat features at the site and reach scale, a direct relationship with macroinvertebrate assemblages was only apparent between the extent of walled banks and the secondary PCA axis described by species data. The strong relationship between catchment scale variables and descriptors of physical structure at the reach and site scale suggests that catchment-scale parameters are valuable predicators of macroinvertebrate community structure in these streams despite the anthropogenic modifications of the natural habitat.

Study of tRNA modifying enzymes and codon usage bias in cancer

Recent evidences indicate that tRNA modifications and tRNA modifying enzymes may play important roles in complex human diseases such as cancer, neurological disorders and mitochondrial-linked diseases. We postulate that expression deregulation of tRNA modifying enzymes affects the level of tRNA modifications and, consequently, their function and the translation efficiency of their tRNA corresponding codons. Due to the degeneracy of the genetic code, most amino acids are encoded by two to six synonymous codons. This degeneracy and the biased usage of synonymous codons cause alterations that can span from protein folding to enhanced translation efficiency of a select gene group. In this work, we focused on cancer and performed a meta-analysis study to compare microarray gene expression profiles, reported by previous studies and evaluate the codon usage of different types of cancer where tRNA modifying enzymes were found de-regulated. A total of 36 different tRNA modifying enzymes were found de-regulated in most cancer datasets analyzed. The codon usage analysis revealed a preference for codons ending in AU for the up-regulated genes, while the down-regulated genes show a preference for GC ending codons. Furthermore, a PCA biplot analysis showed this same tendency. We also analyzed the codon usage of the datasets where the CTU2 tRNA modifying enzyme was found deregulated as this enzyme affects the wobble position (position 34) of specific tRNAs. Our data points to a distinct codon usage pattern between up and downregulated genes in cancer, which might be caused by the deregulation of specific tRNA modifying enzymes. This codon usage bias may augment the transcription and translation efficiency of some genes that otherwise, in a normal situation, would be translated less efficiently.