Search NASA⌕ Search
Home/Archive/DOE OSTI

DOE OSTI · 3373043

From microbial diversity to functional potential using dimensionality reduction

Chamberlain, Emelia J. [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)]·Boulton, William [Univ. of East Anglia, Norwich (United Kingdom)]·Connors, Elizabeth [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)]·Calianos, Theodore [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)]·Bowman, Jeff S. [Univ. of California, San Diego, CA (United States). Scripps Inst. of Oceanography]·Creamean, Jessie M. [Colorado State Univ., Fort Collins, CO (United States)]·Mock, Thomas [Univ. of East Anglia, Norwich (United Kingdom)]·Kim, Heather H. [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)]
View original source ↗Read source document ↗Export metadata ↓

Abstract

The high dimensionality of microbial diversity data from ‘omics observations can be reduced using Machine Learning, with many recent studies showcasing ML utility for exploratory ecological feature finding and process prediction. Here, we compare the Self Organizing Map (SOM) dimensionality reduction method to the well-documented sample-based Principal Coordinate Analysis (PCoA) and taxa-based Weighted Gene Correlation Network Analysis (WGCNA) using near daily 16S rRNA gene amplicon sequencing data from the 2019 to 2020 MOSAiC International Arctic Drift Expedition. We then map k-means clustering outputs from each method to available metagenomes, extracting functionally distinct seasonal microbial ecotypes in the surface Arctic Ocean. Our results indicate the SOM method better represented expected seasonal transitions and identified a greater number of metabolically distinct functional groups than the more traditional PCoA ordination. Ultimately, we identified four community ecotypes with distinct taxonomic and functional cut-offs driven by seasonality, water mass, and substrate turnover, highlighting the importance of succession in functional diversity for the central Arctic Ocean. These results reinforce ML dimensionality reduction as a meaningful translator in the mining of historical amplicon datasets to address modern mechanistic questions and potentially provide ’omics informed ecotype diversity to leverage in mechanistic biogeochemical models.

Explore related subjects

Arctic Oceanbacteriaecosystem functionmachine learningmicrobial diversity

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Chamberlain, Emelia J. [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)], Boulton, William [Univ. of East Anglia, Norwich (United Kingdom)], Connors, Elizabeth [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)], Calianos, Theodore [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)], Bowman, Jeff S. [Univ. of California, San Diego, CA (United States). Scripps Inst. of Oceanography], Creamean, Jessie M. [Colorado State Univ., Fort Collins, CO (United States)], Mock, Thomas [Univ. of East Anglia, Norwich (United Kingdom)], Kim, Heather H. [Woods Hole Oceanographic Institution, Woods Hole, MA (United States)]. 2026-05-18. From microbial diversity to functional potential using dimensionality reduction. https://doi.org/10.3389/fmicb.2026.1786397

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related reports

DOE OSTI2025

Insights into the year-round vertical distribution of chlorophyll concentration in high-latitude Arctic Ocean: implications for primary production

Climate-induced rapid changes in the Arctic Ocean, such as decreasing sea ice extent and increasing water temperature, are altering nutrient and light availability, profoundly impacting primary producer growth. However, access to the high-latitude Arctic Ocean is limited, and satellite data are primarily available only during summer, making continuous in-situ data collection challenging. We collected year-round chlorophyll-a (Chl-a) concentration data in high-latitude regions using a mooring system and performed a comparative analysis with reanalysis data. Unlike previous satellite-based studies, which typically rely on surface measurements, we used the annual vertical distribution of Chl-a. These data were applied to the vertically generalized production model to accurately estimate annual primary production. The moored Chl-a concentration data showed that phytoplankton exhibited a typical subsurface chlorophyll maximum (SCM) layer as sea ice retreated in June. Contrary to the gradually deepening SCM distribution predicted by model-based reanalysis data, the SCM layer persisted for approximately 4 months. This indicates that light and nutrient conditions within the SCM layer remained stable, sustaining continuous phytoplankton growth. Annual primary production, reflecting this vertical distribution of Chl-a concentration, was 6.85 gC m −2 yr −1 . This exceeded satellite-based estimates by at least two-fold, highlighting the significant underestimation of primary production by satellite approaches. Estimating primary production while accounting for the vertical distribution of phytoplankton and light is essential for improving ecological models to better understand carbon cycle and food web changes in the Arctic Ocean, with important implications for climate change predictions.

Arctic Ocean↗