Abstract: Microorganisms play pivotal roles in ecosystems, where precise taxonomic identification is fundamental to understanding and harnessing their functions. Conventional microbial classification, however, relies on pure-culture techniques that suffer from prolonged cultivation cycles, low throughput, high costs, and limited resolution. To address these constraints, we developed an integrated platform combining positive dielectrophoresis-activated Raman-activated cell sorting (pDEP-RACS) with deep ResNet. Using the ecologically versatile and taxonomically challenging genus Pantoea as a model, we constructed a reference Ramanome database by pDEP-RACS that consists of 180 000 single-cell Raman spectra (SCRS) from 12 Pantoea species (22 strains) and two phylogenetically related species and established a classification model by ResNet-18. The model achieved optimal performance 96.9% mean accuracy and 97.3% recall for 24 SCRS colony isolates. Cross-batch validation shows the highest reproducibility in nutrient-starved samples, with 87.9% accuracy postpreprocessing with reduced batch effects. For optimal accuracy (97.6% ± 2.0%), classification accuracy plateaus at >1,500 SCRS of Raman detection depth. In synthetic communities, the model shows ≤3.21% absolute abundance error for species identification. For the rice seed microbiome, a good consistency was observed between Raman-derived Pantoea abundance (34.8%) and 16S rRNA sequencing results (45%). This platform enables species- and strain-level classification of Pantoea spp. cultures, acquiring >7,200 SCRS per hour to facilitate rapid identification in both synthetic microbial communities and field-derived samples.

Abstract: Background: Microbial single-cell Raman spectroscopy (SCRS) has emerged as a powerful tool for label-free phenotyping, enabling rapid characterization of microbial diversity, metabolic states, and functional interactions within complex communities. However, high-throughput SCRS datasets often contain spectral anomalies from noise and fluorescence interference, which obscure microbial signatures and hinder accurate classification. Robust algorithms for outlier detection and microbial ramanome analysis remain underdeveloped. Results: Here, we introduce RamEx, an R package specifically designed for high-throughput microbial ramanome analyses with robust quality control and phenotypic classification. At the core of RamEx is the Iterative Convolutional Outlier Detection (ICOD) algorithm, which dynamically detects spectral anomalies without requiring predefined thresholds. Benchmarking on both simulated and real microbial datasets-including pathogenic bacteria, probiotic strains, and yeast fermentation populations-demonstrated that ICOD achieves an F1 score of 0.97 on simulated datasets and 0.74 on real datasets, outperforming existing approaches by at least 19.8%. Beyond anomaly detection, RamEx provides a modular and scalable workflow for microbial phenotype differentiation, taxonomic marker identification, metabolic-associated fingerprinting, and intra-population heterogeneity analysis. It integrates Raman-based species-specific biomarkers, enabling precise classification of microbial communities and facilitating functional trait mapping at the single-cell level. To support large-scale studies, RamEx incorporates C++ acceleration, GPU parallelization, and optimized memory management, enabling the rapid processing of over one million microbial spectra within an hour. Conclusions: By bridging the gap between high-throughput Raman-based microbial phenotyping and computational analysis, RamEx provides a comprehensive toolkit for exploring microbial ecology, metabolic interactions, and antibiotic susceptibility at the single-cell resolution. RamEx is freely available under the MIT license at https://github.com/qibebt-bioinfo/RamEx. Video Abstract.

Abstract: Heterogeneity within clonal cell populations remains a critical bottleneck within bioprocess engineering, notably by undermining bioproduction yields. Efforts to mitigate its impact have, however, been hampered by technological difficulties quantifying metabolism at the single-cell level. Here, we propose a framework based on single-cell biosensor analysis that enables robust characterisation of cell’s metabolic states, leveraging it to detect and isolate isogeneic heterogeneity in response to environmental perturbations and within microbial cell factories. We identify acute and gradual glucose depletion to induce differentiation of metabolically distinct subpopulations and reveal these subpopulations to exhibit differential production capabilities, with lower intracellular pH subpopulations exhibiting enhanced product accumulation within violacein-producing strains but reduced yields within lycopene-producing strains. Lastly, we highlight galactose cultivation as a method to modulate subpopulation dynamics towards higher-producing lycopene phenotypes. Altogether, our research provides insights into subpopulation differentiation and establishes promising avenues for the engineering of more robust and higher-producing strains.

Abstract: Achieving robust nitrogen removal in constructed wetlands (CWs) is often hindered by the inhibitory effects of high pollutant loads on microbial activity. This study validates that influent load regulation is a pivotal strategy for engineering a partial denitrification (PD)-dominant microbial ecosystem, effectively overcoming such limitations. Comparing low-load (LL), medium-load (ML), and high-load (HL) conditions, the LL group achieved superior total nitrogen removal (up to 97.11%), outperforming the ML and HL groups by 9.28% and 16.21%, respectively. The underlying mechanism involves the alleviation of free ammonia (FA) inhibition at lower loads, which permitted the enrichment of crucial PD-related genera like Pseudomonas and Flavobacterium. This engineered community structure facilitated the stable production of nitrite, a key substrate that subsequently fueled downstream processes such as anammox and DAMO, thereby creating a synergistic nitrogen removal network. Consequently, our findings confirm that managing influent load is a critical and practical method for engineering a stable, PD-dominant microbial ecosystem to achieve robust and sustainable nitrogen removal in CWs.

Abstract: Microbiota-driven fermentation is a global biomanufacturing process that often operates under extreme and fluctuating temperatures. To understand how such systems maintain productivity, this study investigated the Chinese fermentation starter high-temperature Daqu (HTD) as a model system. By combining metagenomics and Raman microspectroscopy, the analysis revealed a drastic decoupling between phylogenetic composition and metabolic activity, with only 10–32 % of yeast species detected by sequencing remaining metabolically active under heat stress. Raman-activated cell sorting and culture (RACS-Culture) recovered three yeasts that consistently maintained viability throughout HTD production: Pichia kudriavzevii, Wickerhamomyces anomalus, and Saccharomycopsis fibuligera. Mono-species and synthetic-community fermentation further revealed a sophisticated mechanism of temporal niche partitioning: in the moderate-temperature early and late stages, S. fibuligera and W. anomalus dominated substrate degradation and flavor precursor biosynthesis, respectively. However, as temperatures rose above 45 °C, both species exhibited low metabolic activity and survival rates. In contrast, only P. kudriavzevii sustained robust growth at this elevated temperature. Genomic analysis revealed a remarkable expansion of heat-resistance and cell-clustering–related genes of wos2 and FLO8 in P. kudriavzevii. These genetic characteristics underpin its enhanced viability, which enables the initially low-abundance species to thrive as a primary ethanol producer and ultimately establish numerical dominance. Thus, temporally overlaying single-cell metabolic vitality profiles onto the corresponding metagenomes can unravel novel functional species and reveal their ecological roles in a complex ecosystem.

Abstract: Cellular metabolic state and its heterogeneity are pivotal features that determine fermentation productivity, yet label-free monitoring has generally been difficult. Employing beer fermentation by Saccharomyces pastorianus as a model, we demonstrated that temporal sampling of ramanomes, the collection of spontaneous Single-Cell Raman Spectra (SCRS) from an isogenic population, provides rich insights into the profiles and inter-conversion of both intra- and extra-cellular metabolites. Among 43 extracellular metabolic phenotypes, ramanomes successfully modeled 19 of them, including the extracellular levels of four alcohols, four esters, four amino acids, two acids, and four mono- and di-saccharide substrates, plus the alcohol-to-ester ratio. Moreover, Intra-Ramanome Correlation Analysis (IRCA) revealed potential metabolic interactions in pairs of intracellular metabolites, extracellular metabolites, and medium substrates. Specifically, carbohydrates were the most active intracellular metabolites, while proteins significantly influenced alcohol and ester synthesis on Day 1 of fermentation. Additionally, both alcohols and esters showed negative correlations with extracellular amino acids and acids. The global-IRCN average degree, reflecting metabolic network complexity, increased over time and was positively correlated with extracellular levels of key products such as n-propanol and various esters, while negatively correlated with acetic acid and certain sugars. Therefore, by enabling non-destructive, label-free, and rapid modeling of both intra- and extracellular metabolite levels, ramanomics can find wide applications in process monitoring and control.