A landmark international study led by the UK’s Medicines and Healthcare products Regulatory Authority (MHRA) has exposed significant variability in microbiome research methods across the world, highlighting an urgent need to adopt unified standards for the global harmonisation of gut microbiome research.
The study – recently published in mSystems an open-access (OA) journal from the American Society for Microbiology - involved leading laboratories across multiple countries testing identical samples of gut microbiome bacteria. Results revealed startling inconsistencies, with accuracy measures varying dramatically between laboratories – despite analysing the same samples.
23 laboratories from 11 countries across four continents were asked to use the newly developed World Health Organization (WHO) International Reference Reagents – containing DNA of known bacterial compositions – to determine whether the methodologies were suitable. When they analysed these samples:
- Species identification varied from 63% to 100% accuracy across different methods, meaning that some laboratories failed to detect a third of the bacterial species present in the sample
- False positive rates ranged from 0% to 41%, meaning some laboratories incorrectly identified bacteria that weren’t present in the sample
- Sample diversity estimates varied considerably, with laboratories identifying as few as 12 or as many as 185 different bacterial species in the same sample.
Microbiome research increasingly informs treatments for conditions including inflammatory bowel disease, metabolic syndrome, irritable bowel syndrome and colorectal cancer. However, contradictory findings between studies have undermined confidence in this rapidly expanding field.
The study identified that a key source of variation may arise from the application of high throughput sequencing methods applied to characterise gut microbiome samples. Notably, the research demonstrated that appropriate analysis methods could achieve accurate results with far fewer resources than previously thought (depending on the sample) – one laboratory achieved quality standards using approximately 890,000 DNA sequence reads, compared with others using an average of 25.4 million reads.
The study establishes Minimum Quality Criteria for four key reporting measures that laboratories worldwide can now use to validate their microbiome analysis methods. These standards are based on real-world data rather than theoretical predictions.
The WHO International DNA Gut Reference Reagents are now available for laboratories globally to benchmark their performance and optimise their methodologies.
Saba Anwar, lead author of the study and Senior Research and Development Scientist at the MHRA, said: “This benchmark study highlights the true level of variability in microbiome data across the world and across sectors, underscoring the critical need for the use of WHO International DNA Gut Reference Reagents to elevate the quality of data in microbiome research. By pinpointing the sources of bias in existing methodologies using reference reagents, we can accelerate innovation and method optimisation.”
This research was commissioned and funded by the Department of Health and Social Care Policy Research Programme through the MHRA Regulatory Science Research Unit, with additional support from an Innovate UK Small Business Research Initiative Award for ‘Creating Standards for Microbiome Therapies’.
All sequencing data are publicly available on the NCBI Sequence Read Archive under BioProject PRJNA1238537. Reference reagents are available at www.nibsc.org/products.aspx.
The research involved 34 participating institutions, including academic institutions, government laboratories, and industry, across multiple countries including the UK, USA, Netherlands, Germany, Korea, Australia, Mexico and Turkey. The study tested both shotgun sequencing and 16S rRNA gene amplicon sequencing methods.
- Anwar S, Lamaudiere M, Hassall J, et al. DNA reference reagents isolate biases in microbiome profiling: a global multi-lab study. mSystems. Published online October 17, 2025. doi:10.1128/msystems.00466-25