Overview of data preprocessing for machine learning applications in human microbiome research

Although metagenomic sequencing is now the preferred technique to study microbiome-host interactions, analyzing and interpreting microbiome sequencing data presents challenges primarily attributed to the statistical specificities of the data (e.g., sparse, over-dispersed, compositional, inter-variable dependency). This mini review explores preprocessing and transformation methods applied in recent human microbiome studies to address microbiome data analysis challenges. Our results indicate a limited adoption of transformation methods targeting the statistical characteristics of microbiome sequencing data. Instead, there is a prevalent usage of relative and normalization-based transformations that do not specifically account for the specific attributes of microbiome data. The information on preprocessing and transformations applied to the data before analysis was incomplete or missing in many publications, leading to reproducibility concerns, comparability issues, and questionable results. We hope this mini review will provide researchers and newcomers to the field of human microbiome research with an up-to-date point of reference for various data transformation tools and assist them in choosing the most suitable transformation method based on their research questions, objectives, and data characteristics.

Compositional functional regression and isotemporal substitution analysis: Methods and application in time-use epidemiology

The distribution of time that people spend in physical activity of various intensities has important health implications. Physical activity (commonly categorised by the intensity into light, moderate and vigorous physical activity), sedentary behaviour and sleep, should not be analysed separately, because they are parts of a time-use composition with a natural constraint of 24 h/day. To find out how are relative reallocations of time between physical activity of various intensities associated with health, herewith we describe compositional scalar-on-function regression and a newly developed compositional functional isotemporal substitution analysis. Physical activity intensity data can be considered as probability density functions, which better reflects the continuous character of their measurement using accelerometers. These probability density functions are characterised by specific properties, such as scale invariance and relative scale, and they are geometrically represented using Bayes spaces with the Hilbert space structure. This makes possible to process them using standard methods of functional data analysis in the L 2 space, via centred logratio (clr) transformation. The scalar-on-function regression with clr transformation of the explanatory probability density functions and compositional functional isotemporal substitution analysis were applied to a dataset from a cross-sectional study on adiposity conducted among school-aged children in the Czech Republic. Theoretical reallocations of time to physical activity of higher intensities were found to be associated with larger and more progressive expected decreases in adiposity. We obtained a detailed insight into the dose-response relationship between physical activity intensity and adiposity, which was enabled by using the compositional functional approach.

The Goldilocks Day for healthy adiposity measures among children and adolescents

Background

The optimal balance of time spent on daily movement behaviors ("The Goldilocks Day") associated with childhood obesity remains unknown.

Objective

To estimate the optimal durations of sleep, sedentary behavior (SB), light physical activity (LPA), and moderate-to-vigorous physical activity (MPVA) associated with excess adiposity in a paediatric population.

Methods

Accelerometer-measured 24-h movement behaviors were obtained from 659 Czech children and adolescents (8-18-year-olds). Adiposity indicators were body mass index z-score, fat mass percentage, fat-free mass index, and visceral adipose tissue. Excess adiposity was defined as exceeding the 85th percentile for an adiposity indicator. Compositional regression analyses were used investigate the associations between movement behaviors and adiposity indicators and estimating "The Goldilocks Day."

Results

The movement behavior composition was associated with visceral adipose tissue (Fdf1 = 3,df2 = 317 = 3.672, p = 0.013) and fat mass percentage (Fdf1 = 3,df2 = 289 = 2.733, p = 0.044) among children and adolescents. The Goldilocks Day consisted of 8.5 h of sleep, 10.8 h of SB, 3.9 h of LPA, and 0.8 h of MVPA among children and 7.5 h of sleep, 12.4 h of SB, 3.6 h of LPA, and 0.5 h of MVPA among adolescents.

Conclusion

Optimizing the time spent sleeping, and in sedentary and physical activities appears to be important in the prevention of excess adiposity.

Advancing microbiome research with machine learning: key findings from the ML4Microbiome COST action

The rapid development of machine learning (ML) techniques has opened up the data-dense field of microbiome research for novel therapeutic, diagnostic, and prognostic applications targeting a wide range of disorders, which could substantially improve healthcare practices in the era of precision medicine. However, several challenges must be addressed to exploit the benefits of ML in this field fully. In particular, there is a need to establish "gold standard" protocols for conducting ML analysis experiments and improve interactions between microbiome researchers and ML experts. The Machine Learning Techniques in Human Microbiome Studies (ML4Microbiome) COST Action CA18131 is a European network established in 2019 to promote collaboration between discovery-oriented microbiome researchers and data-driven ML experts to optimize and standardize ML approaches for microbiome analysis. This perspective paper presents the key achievements of ML4Microbiome, which include identifying predictive and discriminatory 'omics' features, improving repeatability and comparability, developing automation procedures, and defining priority areas for the novel development of ML methods targeting the microbiome. The insights gained from ML4Microbiome will help to maximize the potential of ML in microbiome research and pave the way for new and improved healthcare practices.

Longitudinal reallocations of time between 24-h movement behaviours and their associations with inflammation in children and adolescents: the UP&DOWN study

BACKGROUND

While there is evidence that physical activity, sedentary behaviour (SB) and sleep may all be associated with modified levels of inflammatory markers in adolescents and children, associations with one movement behaviour have not always been adjusted for other movement behaviours, and few studies have considered all movement behaviours in the 24-hour day as an exposure.

PURPOSE

The aim of the study was to explore how longitudinal reallocations of time between moderate-to-vigorous physical activity (MVPA), light physical activity (LPA), SB and sleep are associated with changes in inflammatory markers in children and adolescents.

METHODS

A total of 296 children/adolescents participated in a prospective cohort study with a 3-year follow-up. MVPA, LPA and SB were assessed by accelerometers. Sleep duration was assessed using the Health Behavior in School-aged Children questionnaire. Longitudinal compositional regression models were used to explore how reallocations of time between movement behaviours are associated with changes in inflammatory markers.

RESULTS

Reallocations of time from SB to sleep were associated with increases in C3 levels (difference for 60 min/d reallocation [d₆₀] = 5.29 mg/dl; 95% confidence interval [CI] = 0.28, 10.29) and TNF-α (d₆₀ = 1.81 mg/dl; 95% CI = 0.79, 15.41) levels. Reallocations from LPA to sleep were also associated with increases in C3 levels (d₆₀ = 8.10 mg/dl; 95% CI = 0.79, 15.41). Reallocations from LPA to any of the remaining time-use components were associated with increases in C4 levels (d₆₀ ranging from 2.54 to 3.63 mg/dl; p < 0.05), while any reallocation of time away from MVPA was associated with unfavourable changes in leptin (d₆₀ ranging from 3088.44 to 3448.07 pg/ml; p < 0.05).

CONCLUSIONS

Reallocations of time between 24-h movement behaviours are prospectively associated with some inflammatory markers. Reallocating time away from LPA appears to be most consistently unfavourably associated with inflammatory markers. Given that higher levels of inflammation during childhood and adolescence are associated with an increased risk of chronic diseases in adulthood, children and adolescents should be encouraged to maintain or increase the level of LPA to preserve a healthy immune system.

The Use of Time Flow Analysis to Describe Changes in Physical Ergonomic Work Behaviours Following a Cluster-Randomized Controlled Participatory Ergonomic Intervention

AIM

Evaluations of participatory ergonomic interventions are often challenging as these types of interventions are tailored to the context and need of the workplace in which they are implemented. We aimed to describe how time flow analysis can be used to describe changes in work behaviours following a participatory ergonomic intervention.

METHOD

This study was based on data from a two-arm cluster-randomized controlled trial with 29 childcare institutions and 116 workers (intervention: n = 60, control: n = 56). Physical behaviours at work were technically measured at baseline and 4-month follow-up. Physical behaviours were expressed in terms of relative work time spent forward bending of the back ≥30°, kneeling, active (i.e. walking, stair climbing and running) and sedentary. Average time flow from baseline to follow-up were calculated for both groups to investigate if work time was allocated differently at follow-up.

RESULTS

A total of 116 workers (60 in the intervention and 56 in the control group) had valid accelerometer at baseline and follow-up. The largest group difference in time flowing from baseline to follow-up was observed for forward bending of the back and time spent kneeling. Compared to the control, the intervention group had less time flowing from forward bending of the back to kneeling (intervention: +11 min day, control: +16 min day) and more time flowing from kneeling to sedentary behaviours (intervention: +15 min day, control: +10 min day).

CONCLUSION

The results of this study showed that time flow analysis can be used to reveal changes in work time-use following a participatory ergonomic intervention. For example, the analysis revealed that the intervention group had replaced more work time spent kneeling with sedentary behaviours compared to the control group. This type of information on group differences in time reallocations would not have been possible to obtain by comparing group differences in work time-use following the intervention, supporting the usefulness of time flow analysis as a tool to evaluate complex, context-specific interventions.

Compositional cubes: a new concept for multi-factorial compositions

Compositional data are commonly known as multivariate observations carrying relative information. Even though the case of vector or even two-factorial compositional data (compositional tables) is already well described in the literature, there is still a need for a comprehensive approach to the analysis of multi-factorial relative-valued data. Therefore, this contribution builds around the current knowledge about compositional data a general theoretical framework for k-factorial compositional data. As a main finding it turns out that, similar to the case of compositional tables, also the multi-factorial structures can be orthogonally decomposed into an independent and several interactive parts and, moreover, a coordinate representation allowing for their separate analysis by standard analytical methods can be constructed. For the sake of simplicity, these features are explained in detail for the case of three-factorial compositions (compositional cubes), followed by an outline covering the general case. The three-dimensional structure is analyzed in depth in two practical examples, dealing with systems of spatial and time dependent compositional cubes. The methodology is implemented in the R package robCompositions.

Changes in sedentary behavior patterns during the transition from childhood to adolescence and their association with adiposity: a prospective study based on compositional data analysis

Background

To date, no longitudinal study using a compositional approach has examined sedentary behavior (SB) patterns in relation to adiposity in the pediatric population. Therefore, our aims were to (1) investigate the changes in SB patterns and adiposity from childhood to adolescence, (2) analyze the prospective compositional associations between changes in SB patterns and adiposity, and (3) estimate the changes in adiposity associated with substituting SB with physical activity (PA) of different intensities.

Methods

The study presents a longitudinal design with a 5-year follow-up. A total of 88 participants (61% girls) were included in the analysis. PA and SB were monitored for seven consecutive days using a hip-worn accelerometer. Adiposity markers (fat mass percentage [FM%], fat mass index [FMI], and visceral adiposity tissue [VAT]) were assessed using the multi-frequency bioimpedance analysis. The prospective associations were examined using compositional data analysis.

Results

Over the follow-up period, the proportion of time spent in total SB increased by 154.8 min/day (p < 0.001). The increase in total SB was caused mainly by an increase in middle and long sedentary bouts, as these SB periods increased by 79.8 min/day and 62.0 min/day (p < 0.001 for both), respectively. FM%, FMI, and VAT increased by 2.4% points, 1.0 kg/m², and 31.5 cm² (p < 0.001 for all), respectively. Relative to the remaining movement behaviors, the increase in time spent in middle sedentary bouts was significantly associated with higher FM% (β_ilr1 = 0.27, 95% confidence interval [CI]: 0.02 to 0.53) at follow-up. Lower VAT by 3.3% (95% CI: 0.8 to 5.7), 3.8% (95% CI: 0.03 to 7.4), 3.9% (95% CI: 0.8 to 6.9), and 3.8% (95% CI: 0.7 to 6.9) was associated with substituting 15 min/week spent in total SB and in short, middle, and long sedentary bouts, respectively, with an equivalent amount of time spent in vigorous PA.

Conclusions

This study showed unfavorable changes in SB patterns and adiposity status in the transition from childhood to adolescence. Incorporating high-intensity PA at the expense of SB appears to be an appropriate approach to reduce the risk of excess adiposity in the pediatric population.

Supplementary Information

The online version contains supplementary material available at 10.1186/s13690-021-00755-5.

Day-to-day pattern of work and leisure time physical behaviours: are low socioeconomic status adults couch potatoes or work warriors?

Background

Most studies on day-to-day patterns of physical behaviours (i.e. physical activities and sedentary behaviour) are based on adults with high socioeconomic status (SES) and without differentiating between work and leisure time. Thus, we aimed to characterise the day-to-day leisure time physical behaviours patterns among low SES adults and investigate the influence of work physical behaviours.

Methods

This cross-sectional study included 963 adults from low SES occupations (e.g. manufacturing, cleaning and transportation). The participants wore accelerometers for 1–7 days to measure physical behaviours during work and leisure time, expressed as time-use compositions consisting of time spent sedentary, standing or being active (walking, running, stair climbing, or cycling). Compositional multivariate multilevel models were used to regress daily leisure time-use composition against work time-use compositions. Interaction between weekday and (1) type of day, (i.e., work/non-work) and (2) the work time-use composition were tested. Compositional isotemporal substitution was used to interpret the estimates from the models.

Results

Each weekday, workers consistently spent most leisure time being sedentary and most work time standing. Leisure time physical behaviours were associated with type of day (p < 0.005, more sedentary on workdays vs. non-workdays), weekday (p < 0.005, more sedentary on Friday, Saturday and Sunday), standing work (p < 0.005, more sedentary and less standing and active leisure time on Sunday), and active work (p < 0.005, less sedentary and more standing and active leisure time on Sunday). Sedentary leisure time increased by 18 min, while standing and active leisure time decreased by 11 and 7 min, respectively, when 30 min were reallocated to standing at work on Sunday. Conversely, sedentary leisure time decreased by 25 min, and standing and active leisure time increased by 15 and 10 min, respectively, when 30 min were reallocated to active time at work on Sunday.

Conclusions

While low SES adults’ leisure time was mostly sedentary, their work time was predominantly standing. Work physical behaviours differently influenced day-to-day leisure time behaviours. Thus, public health initiatives aiming to change leisure time behaviours among low SES adults should consider the influence of work physical behaviours.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12889-021-11409-0.

A study on prospective associations between adiposity and 7-year changes in movement behaviors among older women based on compositional data analysis

INTRODUCTION

It is unclear whether adiposity leads to changes in movement behaviors, and there is a lack of compositional analyses of longitudinal data which focus on these associations. Using a compositional approach, this study aimed to examine the associations between baseline adiposity and 7-year changes in physical activity (PA) and sedentary behavior (SB) among elderly women. We also explored the longitudinal associations between change in adiposity and change in movement-behavior composition.

METHODS

This longitudinal study included 176 older women (mean baseline age 62.8 (4.1) years) from Central Europe. Movement behavior was assessed by accelerometers and adiposity was measured by bioelectrical impedance analysis at baseline and follow-up. A set of multivariate least-squares regression analyses was used to examine the associations of baseline adiposity and longitudinal changes in adiposity as explanatory variables with longitudinal changes in a 3-part movement-behavior composition consisting of SB, light PA (LPA) and moderate-to-vigorous PA (MVPA) as outcome variables.

RESULTS

No significant associations were found between baseline adiposity and longitudinal changes in the movement-behavior composition (p > 0.05). We found significant associations of changes in body mass index (BMI) and fat mass percentage (FM%) with changes in the movement-behavior composition. An increase in BMI was associated with an increase of SB at the expense of LPA and MVPA (β = 0.042, p = 0.009) and with a decrease of MVPA in favor of SB and LPA (β = - 0.059, p = 0.037). An increase in FM% was significantly associated only with an increase of SB at the expense of LPA and MVPA (β = 0.019, p = 0.031).

CONCLUSIONS

This study did not support the assumption that baseline adiposity is associated with longitudinal changes in movement behaviors among elderly women, but we found evidence for change-to-change associations, suggesting that a 7-year increase in adiposity is associated with a concurrent increase of SB at the expense of LPA and MVPA and with a concurrent decrease of MVPA in favor of LPA and SB. Public health interventions are needed to simultaneously prevent weight gain and promote physically active lifestyle among elderly women.

Statistical and Machine Learning Techniques in Human Microbiome Studies: Contemporary Challenges and Solutions

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.

Applications of Machine Learning in Human Microbiome Studies: A Review on Feature Selection, Biomarker Identification, Disease Prediction and Treatment

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.

How do short sleepers use extra waking hours? A compositional analysis of 24-h time-use patterns among children and adolescents

BACKGROUND

To examine compositional associations between short sleep duration and sedentary behavior (SB), light physical activity (LPA) and moderate-to-vigorous physical activity (MVPA) among children and adolescents.

METHODS

Multi-day 24-h data on sleep, SB, LPA and MVPA were collected using accelerometers among 343 children (8-13 years old) and 316 adolescents (14-18 years old). Children and adolescents with sleep duration of < 9 and < 8 h, respectively, were classified as short sleepers. Robust compositional regression analysis was used to examine the associations between short sleep duration and the waking-time composition.

RESULTS

Seventy-one percent of children and 75.3% of adolescents were classified as short sleepers. In children, being a short sleeper was associated with higher SB by 95 min/day (p < 0.001) and lower MVPA by 16 min/day (p = 0.002). Specifically, it was associated with a higher amount of time spent in long sedentary bouts (β_ilr1 = 0.46, 95% confidence interval [CI] = 0.29 to 0.62) and lower amounts of time spent in sporadic SB (β_ilr1 = - 0.17, 95% CI = -0.24 to - 0.10), sporadic LPA (β_ilr1 = - 0.09, 95% CI = -0.14 to - 0.04) and sporadic MVPA (β_ilr1 = - 0.17, 95% CI = -0.25 to - 0.10, p < 0.001 for all), relative to the remaining behaviours. In adolescents, being a short sleeper was associated with a higher amount of time spent in SB by 67 min/day (p = 0.001) and lower LPA by 2 min/day (p = 0.035). Specifically, it was associated with more time spent in sedentary bouts of 1-9 min (β_ilr1 = 0.08, 95% CI = 0.02 to 0.14, p = 0.007) and 10-29 min (β_ilr1 = 0.10, 95% CI = 0.02 to 0.18, p = 0.015), relative to the remaining behaviours.

CONCLUSIONS

Among children and adolescents, short sleep duration seems to be highly prevalent and associated with less healthy waking time. Public health interventions and strategies to tackle the high prevalence of short sleep duration among children and adolescents are warranted.

A comparison of seed germination coefficients using functional regression

PREMISE

Seed germination over time is characterized by a sigmoid curve, called a germination curve, in which the percentage (or absolute number) of seeds that have completed germination is plotted against time. A number of individual coefficients have been developed to characterize this germination curve. However, as germination is considered to be a qualitative developmental response of an individual seed that occurs at one time point, but individual seeds within a given treatment respond at different time points, it has proven difficult to develop a single index that satisfactorily incorporates both percentage and rate. The aim of this paper is to develop a new coefficient, the continuous germination index (CGI), which quantifies seed germination as a continuous process, and to compare the CGI with other commonly used indexes.

METHODS

To create the new index, the germination curves were smoothed using nondecreasing splines and the CGI was derived as the area under the resulting spline. For the comparison of the CGI with other common indexes, a regression model with functional response was developed.

RESULTS

Using both an experimentally obtained wild pea (Pisum sativum subsp. elatius) seed data set and a hypothetical data set, we showed that the CGI is able to characterize the germination process better than most other indices. The CGI captures the local behavior of the germination curves particularly well.

DISCUSSION

The CGI can be used advantageously for the characterization of the germination process. Moreover, B-spline coefficients extracted by its construction can be employed for the further statistical processing of germination curves using functional data analysis methods.

Cox regression survival analysis with compositional covariates: Application to modelling mortality risk from 24-h physical activity patterns

Survival analysis is commonly conducted in medical and public health research to assess the association of an exposure or intervention with a hard end outcome such as mortality. The Cox (proportional hazards) regression model is probably the most popular statistical tool used in this context. However, when the exposure includes compositional covariables (that is, variables representing a relative makeup such as a nutritional or physical activity behaviour composition), some basic assumptions of the Cox regression model and associated significance tests are violated. Compositional variables involve an intrinsic interplay between one another which precludes results and conclusions based on considering them in isolation as is ordinarily done. In this work, we introduce a formulation of the Cox regression model in terms of log-ratio coordinates which suitably deals with the constraints of compositional covariates, facilitates the use of common statistical inference methods, and allows for scientifically meaningful interpretations. We illustrate its practical application to a public health problem: the estimation of the mortality hazard associated with the composition of daily activity behaviour (physical activity, sitting time and sleep) using data from the U.S. National Health and Nutrition Examination Survey (NHANES).

Physical Dormancy Release in Medicago truncatula Seeds Is Related to Environmental Variations

Seed dormancy and timing of its release is an important developmental transition determining the survival of individuals, populations, and species in variable environments. Medicago truncatula was used as a model to study physical seed dormancy at the ecological and genetics level. The effect of alternating temperatures, as one of the causes releasing physical seed dormancy, was tested in 178 M. truncatula accessions over three years. Several coefficients of dormancy release were related to environmental variables. Dormancy varied greatly (4-100%) across accessions as well as year of experiment. We observed overall higher physical dormancy release under more alternating temperatures (35/15 °C) in comparison with less alternating ones (25/15 °C). Accessions from more arid climates released dormancy under higher experimental temperature alternations more than accessions originating from less arid environments. The plasticity of physical dormancy can probably distribute the germination through the year and act as a bet-hedging strategy in arid environments. On the other hand, a slight increase in physical dormancy was observed in accessions from environments with higher among-season temperature variation. Genome-wide association analysis identified 136 candidate genes related to secondary metabolite synthesis, hormone regulation, and modification of the cell wall. The activity of these genes might mediate seed coat permeability and, ultimately, imbibition and germination.

Sedentary behavior patterns and adiposity in children: a study based on compositional data analysis

BACKGROUND

Between-person differences in sedentary patterns should be considered to understand the role of sedentary behavior (SB) in the development of childhood obesity. This study took a novel approach based on compositional data analysis to examine associations between SB patterns and adiposity and investigate differences in adiposity associated with time reallocation between time spent in sedentary bouts of different duration and physical activity.

METHODS

An analysis of cross-sectional data was performed in 425 children aged 7-12 years (58% girls). Waking behaviors were assessed using ActiGraph GT3X accelerometer for seven consecutive days. Multi-frequency bioimpedance measurement was used to determine adiposity. Compositional regression models with robust estimators were used to analyze associations between sedentary patterns and adiposity markers. To examine differences in adiposity associated with time reallocation, we used the compositional isotemporal substitution model.

RESULTS

Significantly higher fat mass percentage (FM%; β_ilr1 = 0.18; 95% CI: 0.01, 0.34; p = 0.040) and visceral adipose tissue (VAT; β_ilr1 = 0.37; 95% CI: 0.03, 0.71; p = 0.034) were associated with time spent in middle sedentary bouts in duration of 10-29 min (relative to remaining behaviors). No significant associations were found for short (< 10 min) and long sedentary bouts (≥30 min). Substituting the time spent in total SB with moderate-to-vigorous physical activity (MVPA) was associated with a decrease in VAT. Substituting 1 h/week of the time spent in middle sedentary bouts with MVPA was associated with 2.9% (95% CI: 1.2, 4.6), 3.4% (95% CI: 1.2, 5.5), and 6.1% (95% CI: 2.9, 9.2) lower FM%, fat mass index, and VAT, respectively. Moreover, substituting 2 h/week of time spent in middle sedentary bouts with short sedentary bouts was associated with 3.5% (95% CI: 0.02, 6.9) lower FM%.

CONCLUSIONS

Our findings suggest that adiposity status could be improved by increasing MVPA at the expense of time spent in middle sedentary bouts. Some benefits to adiposity may also be expected from replacing middle sedentary bouts with short sedentary bouts, that is, by taking standing or activity breaks more often. These findings may help design more effective interventions to prevent and control childhood obesity.

Bayesian multiple hypotheses testing in compositional analysis of untargeted metabolomic data

Clinical metabolomics aims at finding statistically significant differences in metabolic statuses of patient and control groups with the intention of understanding pathobiochemical processes and identification of clinically useful biomarkers of particular diseases. After the raw measurements are integrated and pre-processed as intensities of chromatographic peaks, the differences between controls and patients are evaluated by both univariate and multivariate statistical methods. The traditional univariate approach relies on t-tests (or their nonparametric alternatives) and the results from multiple testing are misleadingly compared merely by p-values using the so-called volcano plot. This paper proposes a Bayesian counterpart to the widespread univariate analysis, taking into account the compositional character of a metabolome. Since each metabolome is a collection of some small-molecule metabolites in a biological material, the relative structure of metabolomic data, which is inherently contained in ratios between metabolites, is of the main interest. Therefore, a proper choice of logratio coordinates is an essential step for any statistical analysis of such data. In addition, a concept of b-values is introduced together with a Bayesian version of the volcano plot incorporating distance levels of the posterior highest density intervals from zero. The theoretical background of the contribution is illustrated using two data sets containing samples of patients suffering from 3-hydroxy-3-methylglutaryl-CoA lyase deficiency and medium-chain acyl-CoA dehydrogenase deficiency. To evaluate the stability of the proposed method as well as the benefits of the compositional approach, two simulations designed to mimic a loss of samples and a systematical measurement error, respectively, are added.

Adiposity and changes in movement-related behaviors in older adult women in the context of the built environment: a protocol for a prospective cohort study

Background

In older adults, sedentary behaviors increase while physical activity decreases over time following the compositional nature of 24-h behaviors. These changes in movement-related behaviors (MRBs) might be associated with unhealthy weight gain and several health comorbidities. However, information is lacking on how obesity influences longitudinal changes in the composition of MRBs in older adults. Furthermore, the moderating effect of the built environment on prospective associations between obesity and MRBs in older adults is not fully understood. Therefore, using an integrated time-use approach, this study aims to identify prospective associations between obesity and MRBs together with an assessment of the moderating effect of the built environment in elderly women.

Methods

The study was designed as a prospective 7-year follow-up study. It is based on two previous cross-sectional studies that enable the use of participant data (women aged 60+ years, n = 409) as a baseline dataset in the current study. All methods designed for 7-year follow-up are based on previous studies. The data collection comprises device-based measurement of MRBs (ActiGraph GT1M accelerometer), objective assessment of body adiposity (multi-frequency bioelectrical impedance analysis), subjective assessment of the built environment (NEWS-A questionnaire), and other possible confounding factors. Time spent in sedentary behavior, light physical activity, and moderate-to-vigorous physical activity will be used as three components in a composition reflecting individual MRBs. In linear multiple compositional regression analysis assessing the prospective association between obesity and MRBs, the 7-year follow-up composition of the three mentioned components represents the dependent variable. The 7-year changes in the percentage of body fat (body adiposity), baseline composition of MRBs, and parameters of the built environment represent regressors.

Discussion

This study will use an integrated time-use approach to explore causality from obesity to device-measured behaviors in older women. The design and respective analysis consider the compositional nature of MRBs data and the potential moderating effects of various factors. A comprehensive assessment of causality may help to develop multilevel interventional models that enhance physical activity in older adults.

A comparison of generalised linear models and compositional models for ordered categorical data

Ordered categorical data occur in many applied fields, such as geochemistry, econometrics, sociology and demography or even transportation research, for example, in the form of results from various questionnaires. There are different possibilities for modelling proportions of individual categories. Generalised linear models (GLMs) are traditionally used for this purpose, but also methods of compositional data analysis (CoDa) can be considered. Here, both approaches are compared in depth. Particularly, different assumptions of the models on variability are highlighted. Advantages and disadvantages of individual models are pointed out. While the CoDa model may be inappropriate when the variability of the compositional coordinates depends on the regressors, for example, due to different total counts on which the coordinates are based, the GLM may underestimate the uncertainty of the predictions considerably in case of large-scale data.

Variation in wild pea (Pisum sativum subsp. elatius) seed dormancy and its relationship to the environment and seed coat traits

BACKGROUND

Seed germination is one of the earliest key events in the plant life cycle. The timing of transition from seed to seedling is an important developmental stage determining the survival of individuals that influences the status of populations and species. Because of wide geographical distribution and occurrence in diverse habitats, wild pea (Pisum sativum subsp. elatius) offers an excellent model to study physical type of seed dormancy in an ecological context. This study addresses the gap in knowledge of association between the seed dormancy, seed properties and environmental factors, experimentally testing oscillating temperature as dormancy release clue.

METHODS

Seeds of 97 pea accessions were subjected to two germination treatments (oscillating temperatures of 25/15 °C and 35/15 °C) over 28 days. Germination pattern was described using B-spline coefficients that aggregate both final germination and germination speed. Relationships between germination pattern and environmental conditions at the site of origin (soil and bioclimatic variables extracted from WorldClim 2.0 and SoilGrids databases) were studied using principal component analysis, redundancy analysis and ecological niche modelling. Seeds were analyzed for the seed coat thickness, seed morphology, weight and content of proanthocyanidins (PA).

RESULTS

Seed total germination ranged from 0% to 100%. Cluster analysis of germination patterns of seeds under two temperature treatments differentiated the accessions into three groups: (1) non-dormant (28 accessions, mean germination of 92%), (2) dormant at both treatments (29 acc., 15%) and (3) responsive to increasing temperature range (41 acc., with germination change from 15 to 80%). Seed coat thickness differed between groups with dormant and responsive accessions having thicker testa (median 138 and 140 µm) than non-dormant ones (median 84 mm). The total PA content showed to be higher in the seed coat of dormant (mean 2.18 mg g-1) than those of non-dormant (mean 1.77 mg g-1) and responsive accessions (mean 1.87 mg g-1). Each soil and bioclimatic variable and also germination responsivity (representing synthetic variable characterizing germination pattern of seeds) was spatially clustered. However, only one environmental variable (BIO7, i.e., annual temperature range) was significantly related to germination responsivity. Non-dormant and responsive accessions covered almost whole range of BIO7 while dormant accessions are found in the environment with higher annual temperature, smaller temperature variation, seasonality and milder winter. Ecological niche modelling showed a more localized potential distribution of dormant group. Seed dormancy in the wild pea might be part of a bet-hedging mechanism for areas of the Mediterranean basin with more unpredictable water availability in an otherwise seasonal environment. This study provides the framework for analysis of environmental aspects of physical seed dormancy.

Variation in wild pea (Pisum sativum subsp. elatius) seed dormancy and its relationship to the environment and seed coat traits

BACKGROUND

Seed germination is one of the earliest key events in the plant life cycle. The timing of transition from seed to seedling is an important developmental stage determining the survival of individuals that influences the status of populations and species. Because of wide geographical distribution and occurrence in diverse habitats, wild pea (Pisum sativum subsp. elatius) offers an excellent model to study physical type of seed dormancy in an ecological context. This study addresses the gap in knowledge of association between the seed dormancy, seed properties and environmental factors, experimentally testing oscillating temperature as dormancy release clue.

METHODS

Seeds of 97 pea accessions were subjected to two germination treatments (oscillating temperatures of 25/15 °C and 35/15 °C) over 28 days. Germination pattern was described using B-spline coefficients that aggregate both final germination and germination speed. Relationships between germination pattern and environmental conditions at the site of origin (soil and bioclimatic variables extracted from WorldClim 2.0 and SoilGrids databases) were studied using principal component analysis, redundancy analysis and ecological niche modelling. Seeds were analyzed for the seed coat thickness, seed morphology, weight and content of proanthocyanidins (PA).

RESULTS

Seed total germination ranged from 0% to 100%. Cluster analysis of germination patterns of seeds under two temperature treatments differentiated the accessions into three groups: (1) non-dormant (28 accessions, mean germination of 92%), (2) dormant at both treatments (29 acc., 15%) and (3) responsive to increasing temperature range (41 acc., with germination change from 15 to 80%). Seed coat thickness differed between groups with dormant and responsive accessions having thicker testa (median 138 and 140 µm) than non-dormant ones (median 84 mm). The total PA content showed to be higher in the seed coat of dormant (mean 2.18 mg g^-1) than those of non-dormant (mean 1.77 mg g^-1) and responsive accessions (mean 1.87 mg g^-1). Each soil and bioclimatic variable and also germination responsivity (representing synthetic variable characterizing germination pattern of seeds) was spatially clustered. However, only one environmental variable (BIO7, i.e., annual temperature range) was significantly related to germination responsivity. Non-dormant and responsive accessions covered almost whole range of BIO7 while dormant accessions are found in the environment with higher annual temperature, smaller temperature variation, seasonality and milder winter. Ecological niche modelling showed a more localized potential distribution of dormant group. Seed dormancy in the wild pea might be part of a bet-hedging mechanism for areas of the Mediterranean basin with more unpredictable water availability in an otherwise seasonal environment. This study provides the framework for analysis of environmental aspects of physical seed dormancy.

Interpretation of Compositional Regression with Application to Time Budget Analysis

Regression with compositional response or covariates, or even regression between parts of a composition, is frequently employed in social sciences. Among other possible applications, it may help to reveal interesting features in time allocation analysis. As individual activities represent relative contributions to the total amount of time, statistical processing of raw data (frequently represented directly as proportions or percentages) using standard methods may lead to biased results. Specific geometrical features of time budget variables are captured by the logratio methodology of compositional data, whose aim is to build (preferably orthonormal) coordinates to be applied with popular statistical methods. The aim of this paper is to present recent tools of regression analysis within the logratio methodology and apply them to reveal potential relationships among psychometric indicators in a real-world data set. In particular, orthogonal logratio coordinates have been introduced to enhance the interpretability of coefficients in regression models.

A new method for correlation analysis of compositional (environmental) data - a worked example

Most data in environmental sciences and geochemistry are compositional. Already the unit used to report the data (e.g., μg/l, mg/kg, wt%) implies that the analytical results for each element are not free to vary independently of the other measured variables. This is often neglected in statistical analysis, where a simple log-transformation of the single variables is insufficient to put the data into an acceptable geometry. This is also important for bivariate data analysis and for correlation analysis, for which the data need to be appropriately log-ratio transformed. A new approach based on the isometric log-ratio (ilr) transformation, leading to so-called symmetric coordinates, is presented here. Summarizing the correlations in a heat-map gives a powerful tool for bivariate data analysis. Here an application of the new method using a data set from a regional geochemical mapping project based on soil O and C horizon samples is demonstrated. Differences to 'classical' correlation analysis based on log-transformed data are highlighted. The fact that some expected strong positive correlations appear and remain unchanged even following a log-ratio transformation has probably led to the misconception that the special nature of compositional data can be ignored when working with trace elements. The example dataset is employed to demonstrate that using 'classical' correlation analysis and plotting XY diagrams, scatterplots, based on the original or simply log-transformed data can easily lead to severe misinterpretations of the relationships between elements.

The compositional isotemporal substitution model: A method for estimating changes in a health outcome for reallocation of time between sleep, physical activity and sedentary behaviour

How people use their time has been linked with their health. For example, spending more time being physically active is known to be beneficial for health, whereas long durations of sitting have been associated with unfavourable health outcomes. Accordingly, public health messages have advocated swapping strategies to promote the reallocation of time between parts of the time-use composition, such as “Move More, Sit Less”, with the aim of achieving optimal distribution of time for health. However, the majority of research underpinning these public health messages has not considered daily time use as a composition, and has ignored the relative nature of time-use data. We present a way of applying compositional data analysis to estimate change in a health outcome when fixed durations of time are reallocated from one part of a particular time-use composition to another, while the remaining parts are kept constant, based on a multiple linear regression model on isometric log ratio coordinates. In an example, we examine the expected differences in Body Mass Index z-scores for reallocations of time between sleep, physical activity and sedentary behaviour.

Compositional data analysis for physical activity, sedentary time and sleep research

The health effects of daily activity behaviours (physical activity, sedentary time and sleep) are widely studied. While previous research has largely examined activity behaviours in isolation, recent studies have adjusted for multiple behaviours. However, the inclusion of all activity behaviours in traditional multivariate analyses has not been possible due to the perfect multicollinearity of 24-h time budget data. The ensuing lack of adjustment for known effects on the outcome undermines the validity of study findings. We describe a statistical approach that enables the inclusion of all daily activity behaviours, based on the principles of compositional data analysis. Using data from the International Study of Childhood Obesity, Lifestyle and the Environment, we demonstrate the application of compositional multiple linear regression to estimate adiposity from children's daily activity behaviours expressed as isometric log-ratio coordinates. We present a novel method for predicting change in a continuous outcome based on relative changes within a composition, and for calculating associated confidence intervals to allow for statistical inference. The compositional data analysis presented overcomes the lack of adjustment that has plagued traditional statistical methods in the field, and provides robust and reliable insights into the health effects of daily activity behaviours.

Compositional analysis of trade flows structure

Statistical analysis of trade flows structure can significantly help to reveal or to confirm important macroeconomic phenomena. Because of relative character of these multivariate observations, application of standard multivariate methods directly to raw data can lead to meaningless results, affected by trade sizes of different countries. As a way out, it is proposed to employ the logratio methodology that is able to capture interesting features through logratios between compositional parts. Particularly, the perturbation operation together with clr coefficients for coordinate representation of compositions seem to be easy to handle and to interpret for the purpose. Popular exploratory tools, principal component analysis and PARAFAC modeling of threeway data, resulting from a long-term study of the export/import structure, are applied in the compositional context for data from OECD and WIOD databases. The results show that the logratio methodology enables to reveal interesting features of world trade flows and thus provides a preferable alternative to existing exploratory tools.

Error Propagation in Isometric Log-ratio Coordinates for Compositional Data: Theoretical and Practical Considerations

Compositional data, as they typically appear in geochemistry in terms of concentrations of chemical elements in soil samples, need to be expressed in log-ratio coordinates before applying the traditional statistical tools if the relative structure of the data is of primary interest. There are different possibilities for this purpose, like centered log-ratio coefficients, or isometric log-ratio coordinates. In both the approaches, geometric means of the compositional parts are involved, and it is unclear how measurement errors or detection limit problems affect their presentation in coordinates. This problem is investigated theoretically by making use of the theory of error propagation. Due to certain limitations of this approach, the effect of error propagation is also studied by means of simulations. This allows to provide recommendations for practitioners on the amount of error and on the expected distortion of the results, depending on the purpose of the analysis.

Covariance Structure of Compositional Tables

Recent experiences with interpretation of orthonormal coordinates in compositionaldata show clearly a necessity of their better understanding in terms of logratios that formthe primary source of information within the logratio methodology. This is even morecrucial in the special case of compositional tables, where both balances and coordinateswith odds ratio interpretation are involved. The aim of the paper is to provide a decompo-sition of covariance structure of orthonormal coordinates in compositional tables in termsof logratio variances that could serve for this purpose. For their better interpretability,the formulas are also accompanied with appropriate comments and graphical illustrations,and implications for the prominent case of 2 2 compositional tables are discussed.

Untargeted metabolomic analysis of urine samples in the diagnosis of some inherited metabolic disorders

BACKGROUND

Metabolomics is becoming an important tool in clinical research and the diagnosis of human diseases. It has been used in the diagnosis of inherited metabolic disorders with pronounced biochemical abnormalities. The aim of this study was to determine if it could be applied in the diagnosis of inherited metabolic disorders (IMDs) with less clear biochemical profiles from urine samples using an untargeted metabolomic approach.

METHODS

A total of 14 control urine samples and 21 samples from infants with cystinuria, maple syrup urine disease, adenylosuccinate lyase deficiency and galactosemia were tested. Samples were analyzed by liquid chromatography on aminopropyl column in aqueous normal phase separation system using gradient elution of acetonitrile/ammonium acetate. Detection was performed by time-of-flight mass spectrometer fitted with electrospray ionisation in positive mode. The data were statistically processed using principal component analysis (PCA), principal component discriminant function analysis (PCA-DFA) and partial least squares (PLS) regression.

RESULTS

All patient samples were first distinguished from controls using unsupervised PCA. Discrimination of the patient samples was then unambiguously verified using supervised PCA-DFA. Known markers of the diseases in question were successfully confirmed and a potential new marker emerged from the PLS regression.

CONCLUSION

This study showed that untargeted metabolomics can be applied in the diagnosis of mild IMDs with less clear biochemical profiles.

Sparse principal balances

Compositional data analysis deals with situations where the relevant information is contained only in the ratios between the measured variables, and not in the reported values. This article focuses on high-dimensional compositional data (in the sense of hundreds or even thousands of variables), as they appear in chemometrics (e.g., mass spectral data), proteomics or genomics. The goal of this contribution is to perform a dimension reduction of such data, where the new directions should allow for interpretability. An approach named principal balances turned out to be successful for low dimensions. Here, the concept of sparse principal component analysis is proposed for constructing principal directions, the so-called sparse principal balances. They are sparse (contain many zeros), build an orthonormal basis in the sample space of the compositional data, are efficient for dimension reduction and are applicable to high-dimensional data.

Bayesian-multiplicative treatment of count zeros in compositional data sets

Compositional count data are discrete vectors representing the numbers of outcomes falling into any of several mutually exclusive categories. Compositional techniques based on the log-ratio methodology are appropriate in those cases where the total sum of the vector elements is not of interest. Such compositional count data sets can contain zero values which are often the result of insufficiently large samples. That is, they refer to unobserved positive values that may have been observed with a larger number of trials or with a different sampling design. Because the log-ratio transformations require data with positive values, any statistical analysis of count compositions must be preceded by a proper replacement of the zeros. A Bayesian-multiplicative treatment has been proposed for addressing this count zero problem in several case studies. This treatment involves the Dirichlet prior distribution as the conjugate distribution of the multinomial distribution and a multiplicative modification of the non-zero values. Different parameterizations of the prior distribution provide different zero replacement results, whose coherence with the vector space structure of the simplex is stated. Their performance is evaluated from both the theoretical and the computational point of view.