How Microbiome Signatures Are Derived

Overview

The Microbiome Signatures Database curates condition-specific microbial patterns—referred to as microbiome signatures—through a structured and reproducible synthesis process that integrates evidence across the scientific literature. This methodology prioritizes clinical relevance, microbial consistency, and actionable insight.

Overview of Methodology

Each microbiome signature is constructed using the Microbial Shift and Realignment Process (MSRP), a proprietary framework that organizes microbial abundance data into coherent patterns of dysbiosis, aligned with potential therapeutic targets.

Systematic Literature Retrieval and Screening

Relevant studies are identified through domain-specific searches across PubMed, Scopus, and Google Scholar using combinations of MeSH terms and taxonomic keywords (e.g., “Lactobacillus AND endometriosis”). Study design filters refine queries to include human clinical trials, murine models, and in vitro experiments, ensuring both disease-specific and microbiome-focused investigations are captured. All retrieved studies are then screened based on taxonomic resolution (family, genus, or species level), consistency of sample types (e.g., fecal, vaginal, oral), and the reproducibility of microbial shifts across independent cohorts.

Data Extraction and Taxonomic Harmonization

Microbial abundance data are manually extracted and harmonized using standardized taxonomic frameworks, primarily the NCBI Taxonomy and the Genome Taxonomy Database (GTDB) to ensure species-level precision. Nomenclatural inconsistencies are corrected, and synonymous taxa are merged to maintain taxonomic integrity across studies. Each microbial taxon is then categorized based on its relative abundance as either enriched in cases, depleted in cases, or exhibiting inconclusive or mixed evidence. Taxa are tagged at the most granular level reported—such as Blautia wexlerae rather than the broader Blautia genus—when such resolution is consistently available.

Pattern Recognition and Signature Consolidation

Once extracted, microbial data are synthesized to identify recurring taxa associated with the condition. For inclusion, a taxon must meet at least one of the following criteria: statistically significant abundance differences reported in two or more independent studies; a high effect size observed in a large or methodologically rigorous study; or mechanistic relevance supported by functional or metabolic pathway data. Each included taxon is annotated with metadata detailing the directionality of its shift (enriched or depleted), the sample type from which it was derived, the level of taxonomic identification (species or genus), and reference sources with corresponding DOIs.

Major Microbial Associations (MMAs) Designation

A subset of taxa are designated as Major Microbial Associations (MMAs), representing microbes believed to play a causal, modulatory, or pathophysiologically significant role in the condition. MMAs are selected based on several criteria, including concordant findings across multiple human studies, the presence of known virulence or metabolic factors relevant to disease pathogenesis, the ability to modulate immune, hormonal, or barrier-related pathways, and associations with metal ions or xenobiotics implicated in the condition’s etiology. These MMAs serve as primary targets for microbiome-targeted interventions (MBTIs).

Validation with MBTI Alignment

Each microbiome signature is validated against microbiome-targeted interventions (MBTIs) when possible. A valid MBTI must demonstrate the ability to reverse the directionality of key taxa—for example, increasing depleted beneficial microbes or reducing enriched pathobionts—while also improving clinical outcomes in disease-relevant models or human trials. This process supports bidirectional validation: the microbiome signature substantiates the effectiveness of the intervention, and the intervention, in turn, reinforces the biological relevance of the signature. Signatures that meet these criteria are designated as dually validated and are given enhanced translational value within the platform.

Versioning and Continuous Review

Microbiome signatures are continuously updated to incorporate new studies, taxonomic reclassifications, and emerging data on intervention efficacy. Each signature is assigned a version number and includes a last updated date along with detailed methodological notes to ensure transparency, traceability, and reproducibility across iterations.

Access to Source Data

Each signature page includes details of the studies incorporated into the analysis, along with hyperlinked references and concise summaries. Additional notes on study design, sample type, and sequencing technology are provided to support context-specific interpretation. Researchers or clinicians are encouraged to submit new evidence for consideration to support ongoing refinement and transparency.