Explore the vast range of titles available.

Federated learning (FL) enables participants to collaboratively train machine and deep learning models while safeguarding data privacy. However, the FL paradigm still has drawbacks that affect its trustworthiness, as malicious participants could launch adversarial attacks against the training process. Previous research has examined the robustness of horizontal FL scenarios under various attacks. However, there is a lack of research evaluating the robustness of decentralized vertical FL and comparing it with horizontal FL architectures affected by adversarial attacks. Therefore, this study proposes three decentralized FL architectures: HoriChain, VertiChain, and VertiComb. These architectures feature different neural networks and training protocols suitable for horizontal and vertical scenarios. Subsequently, a decentralized, privacy-preserving, and federated use case with non-IID data to classify handwritten digits is deployed to assess the performance of the three architectures. Finally, a series of experiments computes and compares the robustness of the proposed architectures when they are affected by different data poisoning methods, including image watermarks and gradient poisoning adversarial attacks. The experiments demonstrate that while specific configurations of both attacks can undermine the classification performance of the architectures, HoriChain is the most robust one.

Nossa dissertação tem o intuito de analisar o recém-instalado Centro Integrado de Operações de Fronteira (CIOF) na cidade de Foz do Iguaçu. Tal dispositivo foi inspirado nos Fusion Centersestadunidenses, a ideia é criar um mecanismo para troca de informação e inteligência, preparando relatórios usando dados de diferentes fontes e utilizando-os para lutar contra o crime e o terrorismo.Ao longo do trabalho, busca-se identificar as relações de poder envolvidas no processo de implementação do CIOF. Para tal, parte-se da seguinte pergunta: em um contexto mais democrático, a implementação do CIOF em Foz do Iguaçu tem relação com o dispositivo de segurança instalado a partir da Usina Hidrelétrica de Itaipu no período da ditadura militar?.Utilizando as teorias e conceitos desenvolvidos pela Escola de Copenhagen, alargados, e recorrendo a genealogia proposta por Foucault como estratégia analítica, nossa dissertação estuda os diferentes processos de (des)securitização da Tríplice Fronteira que culminam na instalação do CIOF. Demonstra-se como, apesar de ser alvo de um processo de securitização mais agudo, principalmente, após os ataques de 11 de setembro nos EUA, o território é representado e encarado como uma área problemática desde o início do século XX, conforme nossa delimitação temporal.Ao analisar a implementação do CIOF na Tríplice Fronteira em 2019, faz-se a conexão com distintos momentos históricos que à primeira vista parecem não ter relação com tal processo. Da mesma maneira, encontra-se similaridades entre práticas de vigilâncias e controle do território adotadas no passado, que evoluem para também serem aplicadas no presente. Durante a investigação, diversos documentos oficiais e cabos diplomáticos obtidos por meio da Lei de Acesso à Informação foram esmiuçados, assim, traçou-se os processos de (des)securitização pelo qual a região passa.Nota-se que o processo político que leva ao aumento da presença estatal na Tríplice Fronteira, por meio de mecanismos de controle e vigilância, é replicado ao longo do século XX e XXI. Concomitantemente, novas percepções de ameaça surgem, o que justifica um novo ciclo de recrudescimento do dispositivo de segurança, com a normalização da supressão de liberdades individuais e coletivas.

Federated learning (FL) allows participants to collaboratively train machine and deep learning models while protecting data privacy. However, the FL paradigm still presents drawbacks affecting its trustworthiness since malicious participants could launch adversarial attacks against the training process. Related work has studied the robustness of horizontal FL scenarios under different attacks. However, there is a lack of work evaluating the robustness of decentralized vertical FL and comparing it with horizontal FL architectures affected by adversarial attacks. Thus, this work proposes three decentralized FL architectures, one for horizontal and two for vertical scenarios, namely HoriChain, VertiChain, and VertiComb. These architectures present different neural networks and training protocols suitable for horizontal and vertical scenarios. Then, a decentralized, privacy-preserving, and federated use case with non-IID data to classify handwritten digits is deployed to evaluate the performance of the three architectures. Finally, a set of experiments computes and compares the robustness of the proposed architectures when they are affected by different data poisoning based on image watermarks and gradient poisoning adversarial attacks. The experiments show that even though particular configurations of both attacks can destroy the classification performance of the architectures, HoriChain is the most robust one.

Background Probiotics are often used to prevent antibiotic-induced low-diversity dysbiosis, however their effect is not yet sufficiently summarized in this regard. We aimed to investigate the effects of concurrent probiotic supplementation on gut microbiome composition during antibiotic therapy. Methods We performed a systematic review and meta-analysis of randomized controlled trials reporting the differences in gut microbiome diversity between patients on antibiotic therapy with and without concomitant probiotic supplementation. The systematic search was performed in three databases (MEDLINE (via PubMed), Embase, and Cochrane Central Register of Controlled Trials (CENTRAL)) without filters on 15 October 2021. A random-effects model was used to estimate pooled mean differences (MD) with 95% confidence intervals (CI). This review was registered on PROSPERO (CRD42021282983). Results Of 11,769 identified articles, 15 were eligible in the systematic review and 5 in the meta-analyses. Quantitative data synthesis for Shannon (MD = 0.23, 95% CI: [(−)0.06–0.51]), Chao1 (MD = 11.59 [(−)18.42–41.60]) and observed OTUs (operational taxonomic unit) (MD = 17.15 [(−)9.43–43.73]) diversity indices revealed no significant difference between probiotic supplemented and control groups. Lacking data prevented meta-analyzing other diversity indices; however, most of the included studies reported no difference in the other reported α- and ß-diversity indices between the groups. Changes in the taxonomic composition varied across the eligible studies but tended to be similar in both groups. However, they showed a potential tendency to restore baseline levels in both groups after 3–8 weeks. This is the first meta-analysis and the most comprehensive review of the topic to date using high quality methods. The limited number of studies and low sample sizes are the main limitations of our study. Moreover, there was high variability across the studies regarding the indication of antibiotic therapy and the type, dose, and duration of antimicrobials and probiotics. Conclusions Our results showed that probiotic supplementation during antibiotic therapy was not found to be influential on gut microbiome diversity indices. Defining appropriate microbiome diversity indices, their standard ranges, and their clinical relevance would be crucial.

Background Forests have been grazed for millennia. Around the world, forest grazing by livestock became a controversial management practice, gradually restricted in many countries over the past 250 years. This was also the case in most Central and Eastern European countries, including Hungary, where forest grazing was a legally prohibited activity between 1961 and 2017. Until the 2010s, ecologists and nature conservationists considered it merely as a historical form of forest use. As a result, there is little contemporary scientific information available about the impact of forest grazing on vegetation and the traditional ecological knowledge associated with it. Our aim was to explore and summarize this type of knowledge held by herders in Hungary. Methods We interviewed 58 knowledgeable herders and participated in forest grazing activities in 43 study locations across the country. The results were analysed qualitatively. Results We revealed a living ecological knowledge tradition and practice of forest grazing in native and non-native forest stands. The impact of livestock grazing on native and non-native forests is not considerably different, in the view of the herders. For both forest types, the greatest impact of grazing was the suppression of the shrub layer, while grazing also increased the dominance and palatability (“tameness”) of the grasses. Livestock could cause significant damage to seedlings during forest grazing, but if done with care, grazing could also be an integral part of forestry management. Conclusions Sustainability of current forest grazing practices depends on the depth of local and traditional knowledge applied and herders’ stewardship. We stress the importance of collaborating with holders of local and traditional knowledge in order to gain a better understanding of the effects of livestock grazing on vegetation in temperate forests.

To correlate the variable clinical features of oestrogen-receptor-positive breast cancer with somatic alterations, we studied pretreatment tumour biopsies accrued from patients in two studies of neoadjuvant aromatase inhibitor therapy by massively parallel sequencing and analysis. Eighteen significantly mutated genes were identified, including five genes ( RUNX1 , CBFB , MYH9 , MLL3 and SF3B1 ) previously linked to haematopoietic disorders. Mutant MAP3K1 was associated with luminal A status, low-grade histology and low proliferation rates, whereas mutant TP53 was associated with the opposite pattern. Moreover, mutant GATA3 correlated with suppression of proliferation upon aromatase inhibitor treatment. Pathway analysis demonstrated that mutations in MAP2K4 , a MAP3K1 substrate, produced similar perturbations as MAP3K1 loss. Distinct phenotypes in oestrogen-receptor-positive breast cancer are associated with specific patterns of somatic mutations that map into cellular pathways linked to tumour biology, but most recurrent mutations are relatively infrequent. Prospective clinical trials based on these findings will require comprehensive genome sequencing. Whole-genome analysis of oestrogen-receptor-positive tumours in patients treated with aromatase inhibitors show that distinct phenotypes are associated with specific patterns of somatic mutations; however, most recurrent mutations are relatively infrequent so prospective clinical trials will require comprehensive sequencing and large study populations. Tumour responsiveness to aromatase inhibitors Elaine Mardis and colleagues use whole-genome sequencing to gain insight into the mutational landscape of tissue samples from patients with oestrogen-receptor-positive (ER+) breast cancer treated with a neoadjuvant aromatase inhibitor. They identify a number of disease-linked mutations that specifically correlate with tumour-cell histology, proliferation rates and response to treatment. Such information could be used to determine which patients will benefit from aromatase-inhibitor therapy.

New information about the benefits and limitations of testing for resistance to human immunodeficiency virus (HIV) type 1 (HIV-1) drugs has emerged. The International AIDS Society-USA convened a panel of physicians and scientists with expertise in antiretroviral drug management, HIV-1 drug resistance, and patient care to provide updated recommendations for HIV-1 resistance testing. Published data and presentations at scientific conferences, as well as strength of the evidence, were considered. Properly used resistance testing can improve virological outcome among HIV-infected individuals. Resistance testing is recommended in cases of acute or recent HIV infection, for certain patients who have been infected as long as 2 years or more prior to initiating therapy, in cases of antiretroviral failure, and during pregnancy. Limitations of resistance testing remain, and more study is needed to refine optimal use and interpretation.

Background Streptococcus mutans is a member of the human oral microbiota and is considered one of the most important cariogenic organisms. Previous studies have suggested an expansion of S. mutans populations about 10,000 years ago with the onset of agriculture, yet direct molecular evidence of its presence from ancient DNA remains sparse. Results Here, we present population genomic analyses of 25 ancient S. mutans genomes (average read depth 0.1X – 387X) recovered from archaeological remains across Eurasia spanning ~ 8,000 years of human evolution. Recombination-corrected phylogenomic analyses using Gubbins show a star-like phylogeny indicative of an early radiation, with the ancient genomes falling within the genomic diversity of modern isolates but restricted to one of the major clades of the phylogeny (D). Analyses of genes encoding present day virulence factors reveals that the presence of the mutanobactin operon involved in oxygen tolerance is restricted to specific subclades (A & B) and absent among the ancient samples. Using the MEGAHIT assembler followed by binning of contigs with CONCOCT, we recover metagenome-assembled genomes (MAG) of 7 high-coverage ancient S. mutans strains, including a 7,500-year-old sample from an early European Neolithic farmer. Pangenome analysis with modern isolates using the anvi’o’s suite revealed the presence of specific functional genes in the ancient isolates, which were lost through time. Conclusions Our study demonstrates that Streptococcus mutans DNA is well preserved in tooth samples from archaeological remains and show that it formed part of the human oral microbiota already before the onset of agriculture, consistent with a radiation and population expansion well before 8,000 years ago.

Multiple sclerosis (MS) is a neuro-inflammatory and neurodegenerative disease that is most prevalent in Northern Europe. Although it is known that inherited risk for MS is located within or in close proximity to immune-related genes, it is unknown when, where and how this genetic risk originated 1 . Here, by using a large ancient genome dataset from the Mesolithic period to the Bronze Age 2 , along with new Medieval and post-Medieval genomes, we show that the genetic risk for MS rose among pastoralists from the Pontic steppe and was brought into Europe by the Yamnaya-related migration approximately 5,000 years ago. We further show that these MS-associated immunogenetic variants underwent positive selection both within the steppe population and later in Europe, probably driven by pathogenic challenges coinciding with changes in diet, lifestyle and population density. This study highlights the critical importance of the Neolithic period and Bronze Age as determinants of modern immune responses and their subsequent effect on the risk of developing MS in a changing environment. Analysis of a large ancient genome dataset shows that genetic risk for multiple sclerosis rose in steppe pastoralists, providing insight into how genetic ancestry from the Neolithic and Bronze Age has shaped modern immune responses.