Commonly used compositional data analysis implementations are not advantageous in microbial differential abundance analyses benchmarked against biological ground truth
Bibliographic record
Abstract
Abstract Previous benchmarking of differential abundance (DA) analysis methods in microbiome studies have employed synthetic data, simulations, and “real data” examples, but to the best of our knowledge, none have yet employed experimental data with known “ground truth” differential abundance. A key debate in the field centers on whether compositional methods are necessary for DA analysis, which is challenging to answer due to the lack of ground truth data. To address this gap, we created the Bioconductor data package MicrobiomeBenchmarkData , featuring three microbiome datasets with established biological ground truths: 1) diverse oral microbiomes from supragingival and subgingival plaques, expected to favor aerobic and anaerobic bacteria, respectively, 2) low-diversity microbiomes from healthy vaginas and bacterial vaginosis, conditions that have been well-characterized through cell culture and microscopy, and 3) a spike-in dataset with constant, known absolute abundances of three bacteria. We benchmarked 17 DA approaches and demonstrated that compositional DA methods are not beneficial but rather lack sensitivity, show increased variability in constant-abundance spike-ins, and, most surprisingly, more frequently produce paradoxical results with DA in the wrong direction for the low-diversity microbiome. Conversely, commonly used methods in microbiome literature, such as LEfSe , the Wilcoxon test, and RNA-seq-derived methods, performed best. We conclude that researchers continue using widely adopted non-parametric or RNA-seq DA methods and that further development of compositional methods includes benchmarking against datasets with known biological ground truth.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.023 | 0.067 |
| Meta-epidemiology (narrow) | 0.002 | 0.001 |
| Meta-epidemiology (broad) | 0.001 | 0.002 |
| Bibliometrics | 0.002 | 0.002 |
| Science and technology studies | 0.001 | 0.002 |
| Scholarly communication | 0.004 | 0.003 |
| Open science | 0.002 | 0.003 |
| Research integrity | 0.002 | 0.003 |
| Insufficient payload (model declined to judge) | 0.002 | 0.002 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".