Flood risk management: What are the main drivers of prevention?

4 p.

On the optimal usage of labelled examples in semi-supervised multi-class classification problems

In recent years, the performance of semi-supervised learning has been theoretically investigated. However, most of this theoretical development has focussed on binary classification problems. In this paper, we take it a step further by extending the work of Castelli and Cover [1] [2] to the multi-class paradigm. Particularly, we consider the key problem in semi-supervised learning of classifying an unseen instance x into one of K different classes, using a training dataset sampled from a mixture density distribution and composed of l labelled records and u unlabelled examples. Even under the assumption of identifiability of the mixture and having infinite unlabelled examples, labelled records are needed to determine the K decision regions. Therefore, in this paper, we first investigate the minimum number of labelled examples needed to accomplish that task. Then, we propose an optimal multi-class learning algorithm which is a generalisation of the optimal procedure proposed in the literature for binary problems. Finally, we make use of this generalisation to study the probability of error when the binary class constraint is relaxed.

Fast and Efficient Neural Conversion of Human Hematopoietic Cells

Neurons obtained directly from human somatic cells hold great promise for disease modeling and drug screening. Available protocols rely on overexpression of transcription factors using integrative vectors and are often slow, complex, and inefficient. We report a fast and efficient approach for generating induced neural cells (iNCs) directly from human hematopoietic cells using Sendai virus. Upon SOX2 and c-MYC expression, CD133-positive cord blood cells rapidly adopt a neuroepithelial morphology and exhibit high expansion capacity. Under defined neurogenic culture conditions, they express mature neuronal markers and fire spontaneous action potentials that can be modulated with neurotransmitters. SOX2 and c-MYC are also sufficient to convert peripheral blood mononuclear cells into iNCs. However, the conversion process is less efficient and resulting iNCs have limited expansion capacity and electrophysiological activity upon differentiation. Our study demonstrates rapid and efficient generation of iNCs from hematopoietic cells while underscoring the impact of target cells on conversion efficiency.

Seroprevalence of Triatoma virus (Dicistroviridae: Cripaviridae) antibodies in Chagas disease patients

Background: Chagas disease is caused by Trypanosoma cruzi, and humans acquire the parasite by exposure to contaminated feces from hematophagous insect vectors known as triatomines. Triatoma virus (TrV) is the sole viral pathogen of triatomines, and is transmitted among insects through the fecal-oral route and, as it happens with T. cruzi, the infected insects release the virus when defecating during or after blood uptake. Methods: In this work, we analysed the occurrence of anti-TrV antibodies in human sera from Chagas disease endemic and non-endemic countries, and developed a mathematical model to estimate the transmission probability of TrV from insects to man, which ranged between 0.00053 and 0.0015. Results: Our results confirm that people with Chagas disease living in Bolivia, Argentina and Mexico have been exposed to TrV, and that TrV is unable to replicate in human hosts. Conclusions: We presented the first experimental evidence of antibodies against TrV structural proteins in human sera.