The number of microbiome-related studies has notably increased the availability of data on human microbiome composition and function. These studies provide the essential material to deeply explore host-microbiome associations and their relation to the development and progression of various complex diseases. Improved data-analytical tools are needed to exploit all information from these biological datasets, taking into account the peculiarities of microbiome data, i.e., compositional, heterogeneous and sparse nature of these datasets. The possibility of predicting host-phenotypes based on taxonomy-informed feature selection to establish an association between microbiome and predict disease states is beneficial for personalized medicine. In this regard, machine learning (ML) provides new insights into the development of models that can be used to predict outputs, such as classification and prediction in microbiology, infer host phenotypes to predict diseases and use microbial communities to stratify patients by their characterization of state-specific microbial signatures. Here we review the state-of-the-art ML methods and respective software applied in human microbiome studies, performed as part of the COST Action ML4Microbiome activities. This scoping review focuses on the application of ML in microbiome studies related to association and clinical use for diagnostics, prognostics, and therapeutics. Although the data presented here is more related to the bacterial community, many algorithms could be applied in general, regardless of the feature type. This literature and software review covering this broad topic is aligned with the scoping review methodology. The manual identification of data sources has been complemented with: (1) automated publication search through digital libraries of the three major publishers using natural language processing (NLP) Toolkit, and (2) an automated identification of relevant software repositories on GitHub and ranking of the related research papers relying on learning to rank approach.

SummaryWe tested whether cortical porosity of the proximal femur measured using StrAx1.0 software provides additional information to areal bone mineral density (aBMD) or Fracture Risk Assessment Tool (FRAX) in differentiating women with and without fracture. Porosity was associated with fracture independent of aBMD and FRAX and identified additional women with fractures than by osteoporosis or FRAX thresholds.IntroductionNeither aBMD nor the FRAX captures cortical porosity, a major determinant of bone strength. We therefore tested whether combining porosity with aBMD or FRAX improves identification of women with fractures.MethodsWe quantified femoral neck (FN) aBMD using dual-energy X-ray absorptiometry, FRAX score, and femoral subtrochanteric cortical porosity using StrAx1.0 software in 211 postmenopausal women aged 54–94 years with nonvertebral fractures and 232 controls in Tromsø, Norway. Odds ratios (ORs) were calculated using logistic regression analysis.ResultsWomen with fractures had lower FN aBMD, higher FRAX score, and higher cortical porosity than controls (all p < 0.001). Each standard deviation higher porosity was associated with fracture independent of FN aBMD (OR 1.39; 95 % confidence interval 1.11–1.74) and FRAX score (OR 1.58; 1.27–1.97) in all women combined. Porosity was also associated with fracture independent of FRAX score in subgroups with normal FN aBMD (OR 1.88; 1.21–2.94), osteopenia (OR 1.40; 1.06–1.85), but not significantly in those with osteoporosis (OR 1.48; 0.68–3.23). Of the 211 fracture cases, only 18 women (9 %) were identified using FN aBMD T-score < −2.5, 45 women (21 %) using FRAX threshold >20 %, whereas porosity >80th percentile identified 61 women (29 %). Porosity identified 26 % additional women with fractures than identified by the osteoporosis threshold and 21 % additional women with fractures than by this FRAX threshold.ConclusionsCortical porosity is a risk factor for fracture independent of aBMD and FRAX and improves identification of women with fracture.

No abstract

The human microbiome has emerged as a central research topic in human biology and biomedicine. Current microbiome studies generate high-throughput omics data across different body sites, populations, and life stages. Many of the challenges in microbiome research are similar to other high-throughput studies, the quantitative analyses need to address the heterogeneity of data, specific statistical properties, and the remarkable variation in microbiome composition across individuals and body sites. This has led to a broad spectrum of statistical and machine learning challenges that range from study design, data processing, and standardization to analysis, modeling, cross-study comparison, prediction, data science ecosystems, and reproducible reporting. Nevertheless, although many statistics and machine learning approaches and tools have been developed, new techniques are needed to deal with emerging applications and the vast heterogeneity of microbiome data. We review and discuss emerging applications of statistical and machine learning techniques in human microbiome studies and introduce the COST Action CA18131 “ML4Microbiome” that brings together microbiome researchers and machine learning experts to address current challenges such as standardization of analysis pipelines for reproducibility of data analysis results, benchmarking, improvement, or development of existing and new tools and ontologies.

PurposeThere are a high number of HIV-infected patients receiving antiretroviral therapy (ART) in the Kathmandu District of Nepal, but information on adherence and factors influencing it are scarce in this population. The present study aimed to estimate ART adherence among HIV-infected patients in the Kathmandu District of Nepal, and to determine the factors associated with ART adherence.Patients and methodsThis study included 316 HIV-infected patients attending three ART centers in the Kathmandu District. Information on sociodemographic characteristics, socioeconomic status, and ART use for the previous 7 days was collected via interview. Participants were considered adherent if they reported taking ≥95% of their ART as prescribed. The association between explanatory variables and ART adherence was measured using logistic regression and reported as odds ratios (OR) with 95% confidence intervals (CI).ResultsMale participants accounted for 64.6% (n=204). Overall ART adherence was 86.7%. ART adherence in men and women were 84.3% and 91.1%, respectively. Age (OR 1.04; 95% CI 1.00–1.09), travel time to ART centers (OR 1.38; 95% CI 1.12–1.71), history of illegal drug use (OR 3.98; 95% CI 1.71–9.24), and adverse effects (OR 4.88; 95% CI 1.09–21.8), were all independently and negatively associated with ART adherence. Use of reminder tools (OR 3.45; 95% CI 1.33–8.91) was independently and positively associated with ART adherence.ConclusionThe observed ART adherence in this study is encouraging. Travel time to ART centers, self-reported adverse effects, illegal drug use, and not using reminder tools were the major determinants of ART adherence. Interventions that take these factors into account could further improve ART adherence.