Stratification of the Gut Microbiota Composition Landscape across the Alzheimer's Disease Continuum in a Turkish Cohort.

Alzheimer's disease (AD) is a heterogeneous disorder that spans a continuum with multiple phases, including preclinical, mild cognitive impairment, and dementia. Unlike for most other chronic diseases, human studies reporting on AD gut microbiota in the literature are very limited. With the scarcity of approved drugs for AD therapies, the rational and precise modulation of gut microbiota composition using diet and other tools is a promising approach to the management of AD. Such an approach could be personalized if an AD continuum can first be deconstructed into multiple strata based on specific microbiota features by using single or multiomics techniques. However, stratification of AD gut microbiota has not been systematically investigated before, leaving an important research gap for gut microbiota-based therapeutic approaches. Here, we analyze 16S rRNA amplicon sequencing of stool samples from 27 patients with mild cognitive impairment, 47 patients with AD, and 51 nondemented control subjects by using tools compatible with the compositional nature of microbiota. To stratify the AD gut microbiota community, we applied four machine learning techniques, including partitioning around the medoid clustering and fitting a probabilistic Dirichlet mixture model, the latent Dirichlet allocation model, and we performed topological data analysis for population-scale microbiome stratification based on the Mapper algorithm. These four distinct techniques all converge on Prevotella and Bacteroides stratification of the gut microbiota across the AD continuum, while some methods provided fine-scale resolution in stratifying the community landscape. Finally, we demonstrate that the signature taxa and neuropsychometric parameters together robustly classify the groups. Our results provide a framework for precision nutrition approaches aiming to modulate the AD gut microbiota. IMPORTANCE The prevalence of AD worldwide is estimated to reach 131 million by 2050. Most disease-modifying treatments and drug trials have failed, due partly to the heterogeneous and complex nature of the disease. Recent studies demonstrated that gut dybiosis can influence normal brain function through the so-called "gut-brain axis." Modulation of the gut microbiota, therefore, has drawn strong interest in the clinic in the management of the disease. However, there is unmet need for microbiota-informed stratification of AD clinical cohorts for intervention studies aiming to modulate the gut microbiota. Our study fills in this gap and draws attention to the need for microbiota stratification as the first step for microbiota-based therapy. We demonstrate that while Prevotella and Bacteroides clusters are the consensus partitions, the newly developed probabilistic methods can provide fine-scale resolution in partitioning the AD gut microbiome landscape.

Classification of motor imagery and execution signals with population-level feature sets: implications for probe design in fNIRS based BCI.

OBJECTIVE: The aim of this study was to introduce a novel methodology for classification of brain hemodynamic responses collected via functional near infrared spectroscopy (fNIRS) during rest, motor imagery (MI) and motor execution (ME) tasks which involves generating population-level training sets. APPROACH: A 48-channel fNIRS system was utilized to obtain hemodynamic signals from the frontal (FC), primary motor (PMC) and somatosensory cortex (SMC) of ten subjects during an experimental paradigm consisting of MI and ME of various right hand movements. Classification accuracies of random forest (RF), support vector machines (SVM), and artificial neural networks (ANN) were computed at the single subject level by training each classifier with subject specific features, and at the group level by training with features from all subjects for ME versus Rest, MI versus Rest and MI versus ME conditions. The performances were also computed for channel data restricted to FC, PMC and SMC regions separately to determine optimal probe location. MAIN RESULTS: RF, SVM and ANN had comparably high classification accuracies for ME versus Rest (%94, %96 and %98 respectively) and for MI versus Rest (%95, %95 and %98 respectively) when fed with group level feature sets. The accuracy performance of each algorithm in localized brain regions were comparable (>%93) to the accuracy performance obtained with whole brain channels (>%94) for both ME versus Rest and MI versus Rest conditions. SIGNIFICANCE: By demonstrating the feasibility of generating a population level training set with a high classification performance for three different classification algorithms, the findings pave the path for removing the necessity to acquire subject specific training data and hold promise for a novel, real-time fNIRS based BCI system design which will be most effective for application to disease populations for whom obtaining data to train a classification algorithm is not possible.