EXPLANA: A user-friendly workflow for EXPLoratory ANAlysis and feature selection in cross-sectional and longitudinal microbiome studies
biorxiv(2024)
Abstract
The potential for disease treatment through gut microbiome modification has contributed to an increase in longitudinal microbiome studies (LMS). Gut microbiome modification can occur through factors such as diet, probiotics, or fecal transplants. Scientific data often motivates researchers to perform exploratory analyses to identify features that relate to a response. However, LMS are challenging to analyze, often leading to lost information and research barriers. LMS analytic challenges include data integration, compositionality, dimensionality reduction, and the need for mixed-effects models for non-independent data. Additionally, LMS can be observational or interventional, and relevant comparisons of interest might differ for these two study types. For example, in an observational study, measurements are made over time and show natural fluctuations in symptoms/measurements, so the baseline measurement might not be a reference point of primary interest; whereas, in an interventional study, the baseline value often coincides with the start of treatment and is a key reference point. Thus, the optimal way to calculate feature changes for each subject over time is dependent on different reference values. To address these challenges, we developed EXPLANA, a data-driven feature-selection workflow that supports numerical and categorical data. We implemented machine-learning models for repeated measures, feature-selection methods, and visualizers explaining how selected features relate to the response. With one script, analysts can build models to select and evaluate important features and obtain an analytic report that textually and graphically summarizes results. EXPLANA had good performance using twenty simulated data models yielding an average area under the curve (AUC) of 0.91 (range: 0.79-1.0; SD = 0.05) and better performance compared to an existing tool (AUC: 0.95 and 0.56; precision: 0.82, and 0.14, respectively). EXPLANA is a flexible, data-driven tool that simplifies LMS analyses and can identify unique features that are predictive of outcomes of interest through a straightforward workflow.
### Competing Interest Statement
The authors have declared no competing interest.
MoreTranslated text
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined