Written by
Jona Health
Read time
Jona’s perspective on the NIST study, what its findings mean for microbiome testing, and why reliability requires looking beyond individual bacteria to the bigger picture.
The NIST study deserves serious attention. Researchers sent equivalent stool samples to seven consumer microbiome-testing services and found substantial differences across providers, plus one major within-provider outlier. That is a useful warning about methodology, quality control, and overconfident interpretation of the microbiome testing industry.
However, it’s important to distinguish what the study demonstrates from the questions it wasn’t designed to answer. Its main analysis compared the types and amounts of individual organisms reported by each provider. The authors explicitly acknowledge that their reference material did not provide an absolute ground truth, so the study could not determine which provider was most accurate. They also found that reproducibility within a consistent workflow was generally good.
This distinction matters because identifying individual microbes is not the ultimate reason most people are interested in microbiome testing. They want to understand what their microbiome may be telling them about their health and what to do about it. Jona is establishing a new standard for gut health analysis that moves beyond individual microbes to identify function and meaningful patterns across the microbiome.
The NIST study compared providers using fundamentally different methods. Some services used 16S sequencing, which targets regions of a single bacterial gene, while others used whole-metagenomic shotgun sequencing. Collection methods, DNA extraction, sequencing depth, bioinformatics, and reporting thresholds also varied substantially. These different approaches should not necessarily be expected to produce identical lists and quantities of individual organisms.
This isn’t unique to the microbiome. Different measurement methods commonly produce somewhat different results in medicine. An MRI, ultrasound and CT of the heart, for example, provide different views and may produce different measurements.
To continue the cardiac imaging analogy, individual pixels may vary between an MRI and CT, or even between scanners from different manufacturers. But the more important question is whether those differences materially affect the higher-level measurement we care about, such as ejection fraction.
We believe the same principle applies to the microbiome.
Jona uses deep shotgun metagenomic sequencing, the most comprehensive sequencing approach, and the gold standard for current microbiome research. Rather than targeting one bacterial gene, shotgun sequencing reads DNA across the microbial community, including genes that encode metabolic pathways. This allows Jona to estimate the microbiome's functional potential: what it may be capable of producing, which functions have greater or lesser encoded capacity, and which metabolic pathways appear enriched or depleted.
Further, many health associations in microbiome research are not based on one organism being "good" or "bad." They are compositional biomarkers: patterns distributed across the entire ecosystem (much like in cardiac imaging), with multiple organisms elevated or depleted together. Jona's AI evaluates these multiorganism patterns against the scientific literature.
This pattern-based approach is fundamental to how Jona analyzes the microbiome. Rather than treating the output as a list of individual microbes to interpret one by one, Jona looks across the data for larger patterns that provide a more meaningful picture of the microbiome and its relationship to health. We believe this represents a new standard for gut health analysis.
This distinction is important when thinking about the variability identified by NIST. Jona’s analysis is designed to look at these larger-scale patterns rather than depend on any single organism measurement. This raises an important hypothesis. Some variation at the individual-organism level may not materially change the broader signal.
The NIST study did not systematically evaluate the reproducibility of shotgun-derived gene and pathway profiles or the multiorganism biomarkers that have been identified in the literature. So while the study raises an important issue about differences in organism-level measurements across testing providers, it does not tell us whether these higher-level signals linked to assessing health demonstrate the same variability.
The study also reinforces several principles that have shaped Jona’s approach: use a consistent, comprehensive sequencing workflow; apply strong quality controls; avoid reducing health to isolated "good" and "bad" microbes; distinguish association from diagnosis; show the evidence behind every interpretation; and remain product-neutral rather than using a stool test used to sell proprietary supplements.
The NIST study raises important questions for the microbiome testing industry and reinforces the need for greater standardization, transparency and validation. But we believe the conversation about standards needs to extend beyond how microbiome data are generated to how these data are analyzed.
How reproducible are the larger patterns that may tell us something meaningful about health, even when measurements of individual organisms vary?
We believe this is a question worth testing directly. One way to do that would be to apply the same Jona analysis to the results generated from the same stool samples across the seven providers in the NIST study and evaluate how consistent those higher-level patterns are.
Our hypothesis is that the larger patterns will prove more consistent than the individual-organism measurements. If that hypothesis holds, it could help establish a more meaningful standard for reliability in microbiome analysis and support a shift from analyzing microbes in isolation to understanding the microbiome as an interconnected ecosystem.
