ConsensusMetaDA: an R Package for consensus-based differential abundance analysis of metagenomic data.

Differential abundance (DA) analysis is critical in metagenomics research for identifying microbial taxa associated with variable biological conditions. Despite numerous mature DA tools being available, they often generate disparate results, making it challenging to obtain consistent and biologically meaningful findings. Tools differ substantially in their input requirements, underlying statistical models, and filtering strategies, yet researchers typically rely on a single tool due to the practical challenges of integrating multiple algorithms with variable workflows and non-standardized output formats. To address this challenge, we developed ConsensusMetaDA, an easy-to-use R package that integrates five widely-used DA tools (ALDEx2, DESeq2, EdgeR, metagenomeSeq, and ADAPT) into a streamlined three-function workflow. The package generates unified output reports and comprehensive visualizations including rarefaction curves, alpha/beta diversity analyses, taxonomic abundance plots, and Venn/UpSet diagrams to visualize algorithmic consensus and disagreement. We benchmarked ConsensusMetaDA on well-characterized public datasets and simulated data, demonstrating that individual tools exhibit complementary strengths and weaknesses. On datasets with known null results, EdgeR and DESeq2 reported false positive differentially abundant OTUs while ALDEx2, metagenomeSeq, and ADAPT correctly identified none. Conversely, simulations revealed that ALDEx2, metagenomeSeq, and ADAPT frequently missed true positives that EdgeR and DESeq2 detected. Our consensus approach improves both sensitivity and specificity compared to any single tool and enables researchers to minimize either false positive or false negative rates based on study-specific priorities. ConsensusMetaDA is freely available with comprehensive documentation at: https://github.com/FieldLabFNQOmics/ConsensusMetaDA