A "Landmark" Skin Microbiome Map Has a Narrower Claim Than It Sounds

A paper going up this month in the Journal of Investigative Dermatology maps 26 bacterial and fungal species across 24 points on the human face and renders the result as a continuous 3D surface. The authors, led by Martin Stalder and Martin P. Pagac, work for dsm-firmenich, the Dutch-Swiss flavor, fragrance and nutrition company formed by the 2023 DSM-Firmenich merger. That authorship matters for how the findings should be read, because this is a company publishing results about its own diagnostic technology in a peer-reviewed clinical journal, not an outside lab testing a commercial claim.

dsm-firmenich announced the same work in a press release dated June 25, 2026, months before the journal version appears, and had already previewed the mapping approach at the IFSCC Congress in 2025 as part of a wider skin-health product portfolio. The JID paper is the peer-reviewed capstone of a project the company had been describing commercially for more than a year. Nearly every trade write-up of the findings, from Personal Care Insights to SpecialChem to Cosmetics Business, traces back to that same press release rather than to an independent reading of the paper, so the specific numbers and framing circulating online are company self-description until checked against the published methods.

The technical claim worth isolating is narrower than the headline suggests. The team measured absolute microbial density, meaning actual counts per site, rather than the relative abundance figures that dominate most microbiome studies and that make cross-sample comparison harder. Combined with Bio-Me, a Norwegian microbiome-diagnostics company that likely supplied the quantification platform, machine learning was then used to interpolate between the 24 sampled sites into a smooth 3D surface rather than a set of discrete points. That interpolation is a visualization method, not a new biological finding on its own.

It is also not the first 3D map of the facial microbiome. A 2022 study in Acta Dermato-Venereologica, "Biogeographical Landscape of the Human Face Skin Microbiome Viewed in High Definition," already built a high-resolution 3D map from 1,649 samples across 12 individuals using 16S rRNA sequencing, covering one side of the face. That earlier work reported relative composition and how much it varied between people. The dsm-firmenich team's novelty claim rests specifically on two narrower additions: converting the map to absolute density and folding fungal species into the same framework alongside bacteria, something the 2022 study did not attempt.

Inside that absolute-density data, the authors flag a pattern of possible competition between Malassezia restricta and Cutibacterium acnes, species already linked separately to dandruff and seborrheic dermatitis on one side and acne on the other. That is the finding with clinical weight, since it suggests the two organisms occupy overlapping territory on the face and may suppress each other locally. But the paper's own scope here is an observed spatial association in the dataset mapped, not a tested mechanism; whether the pattern holds across ages, skin types or seasons, or whether it is causal rather than coincidental co-occurrence, is outside what 24 to 9 sampling sites on a measured cohort can establish.

A separate machine-learning result in the paper suggests that sampling as few as 9 of the 24 facial sites, with the central forehead flagged as the most informative single spot, can reproduce most of the variability seen in the full map. That is a practical finding aimed at making future testing products cheaper and faster, and it reads as a methods result for product development rather than a dermatological one. dsm-firmenich already sells microbiome-targeted actives, including its HMO-based GLYCARE SK line launched in May 2026, and the company frames the new mapping tool explicitly as a way to monitor microbiome-modulating treatments and support claims for interventions aimed at skin dysbiosis. The central forehead, in other words, may end up mattering less as a biological landmark than as the spot a nine-point consumer diagnostic settles on.