Typical disease-associated microbiota changes are widely studied as potential diagnostic or therapeutic targets. Our aim was to analyze a hospitalized cohort including various gastroenterological pathologies in order to fine-map the gut microbiota dysbiosis. Bacterial (V3 V4) and fungal (ITS2) communities were determined in 121 hospitalized gastrointestinal patients from a single ward and compared to 162 healthy controls. Random Forest models implemented in this study indicated that the gut community structure is in most cases not sufficient to differentiate the subjects based on their underlying disease. Instead, hospitalized patients in our study formed three distinct disease non-related clusters (C1, C2, and C3), partially explained by antibiotic use. Majority of the subjects (cluster C1) closely resembled healthy controls, showing only mild signs of community disruption; most significantly decreased in this cluster were Faecalibacterium and Roseburia. The remaining two clusters (C2 and C3) were characterized by severe signs of dysbiosis; cluster C2 was associated with an increase in Enterobacteriaceae and cluster C3 by an increase in Enterococcus. According to the cluster affiliation, subjects also showed different degrees of inflammation, most prominent was the positive correlation between levels of C-reactive protein and the abundance of Enterococcus.
We address the task of multi-target regression, where we generate global models that simultaneously predict multiple continuous variables. We use ensembles of generalized decision trees, called predictive clustering trees (PCTs), in particular bagging and random forests (RF) of PCTs and extremely randomized PCTs (extra PCTs). We add another dimension of randomization to these ensemble methods by learning individual base models that consider random subsets of target variables, while leaving the input space randomizations (in RF PCTs and extra PCTs) intact. Moreover, we propose a new ensemble prediction aggregation function, where the final ensemble prediction for a given target is influenced only by those base models that considered it during learning. An extensive experimental evaluation on a range of benchmark datasets has been conducted, where the extended ensemble methods were compared to the original ensemble methods, individual multi-target regression trees, and ensembles of single-target regression trees in terms of predictive performance, running times and model sizes. The results show that the proposed ensemble extension can yield better predictive performance, reduce learning time or both, without a considerable change in model size. The newly proposed aggregation function gives best results when used with extremely randomized PCTs. We also include a comparison with three competing methods, namely random linear target combinations and two variants of random projections.
Clostridium difficile infection (CDI) is typically associated with disturbed gut microbiota and changes related to decreased colonization resistance against C. difficile are well described. However, nothing is known about possible effects of C. difficile on gut microbiota restoration during or after CDI. In this study, we have mimicked such a situation by using C. difficile conditioned medium of six different C. difficile strains belonging to PCR ribotypes 027 and 014/020 for cultivation of fecal microbiota. A marked decrease of microbial diversity was observed in conditioned medium of both tested ribotypes. The majority of differences occurred within the phylum Firmicutes, with a general decrease of gut commensals with putative protective functions (i.e. Lactobacillus, Clostridium_XIVa) and an increase in opportunistic pathogens (i.e. Enterococcus). Bacterial populations in conditioned medium differed between the two C. difficile ribotypes, 027 and 014/020 and are likely associated with nutrient availability. Fecal microbiota cultivated in medium conditioned by E. coli, Salmonella Enteritidis or Staphylococcus epidermidis grouped together and was clearly different from microbiota cultivated in C. difficile conditioned medium suggesting that C. difficile effects are specific. Our results show that the changes observed in microbiota of CDI patients are partially directly influenced by C. difficile.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.