Flexible biomarker discovery via phylogeny- and function-aware feature extraction

The Microbiome Elastic Feature Extraction (MEFE) algorithm is a method that leverages phylogenetic relationships, sequence similarity, and functional homology to elastically retrieve microbial candidate biomarkers. Evaluation on both synthetic and real microbiome datasets demonstrates that MEFE significantly reduces false positives and negatives compared to widely used methods, thereby improving microbiome status discrimination and disease detection.