Microorganisms associated with insects can offer diverse benefits to their host, but the degree to which hosts rely on them varies across insect groups. Although research has increasingly focused on Lepidoptera, many of which are important pest species, key aspects of their microbiome's stability and functional importance remain ambiguous and seem context-dependent. The caterpillars of the goat moth, Cossus cossus (Linnaeus), which develop over 2 to 5 years within trunks of deciduous trees and consume a wood-based diet rich in recalcitrant lignin, provide an intriguing model to examine potential microbial contributions to lignocellulose degradation. Full-length 16S rRNA gene sequencing revealed a gut microbiome dominated by Enterococcus spp. and a high Leuconostoc spp. presence in the mandibular glands. Concurrent culturomics approaches recovered 64 taxa identified by sequencing and uncovered additional diversity, such as Brevibacterium spp. and Streptomyces spp. Notably, several isolates, including strains of Acinetobacter johnsonii, Citrobacter gillenii, Leucobacter sp., and Pseudomonas sp., demonstrated growth on minimal media supplemented with wood, lignin, or lignin-derived aromatics as the sole carbon and energy sources. These findings provide the first evidence that specific bacterial members of the goat moth caterpillar's microbiome may contribute to wood digestion, suggesting a functional symbiosis suited to a challenging dietary niche.

Honey bees (Apis mellifera) contribute to crop production and floral biodiversity via pollination, but their health is increasingly challenged by stressors including pathogens, parasites and agricultural practices. Although the honey bee gut microbiome is relatively simple, its phages are not well studied. Here, we conducted a metagenomic study, providing a comprehensive catalogue of honey bee gut phages from 450 virus-enriched samples from 63 hives, across eight European countries, three seasons and three gut sections. We describe a diverse phageome including many phages that appear to belong to novel taxa, as well as a core set of 97 highly prevalent phages. In addition, we identify potential auxiliary metabolic genes, such as a sulfur metabolism gene carried by phages that are predominantly temperate and likely infect mutualistic honey bee core bacteria. This gene is associated with land use around the sampled hives, indicating complex ecological interactions in the tripartite system of the honey bee, its microbiota and the phages therein.

The story of Belgian lambic beers goes that they can only be produced around the Senne river since the valley of this river harbors essential yeasts that inoculate the lambic beer wort during its cooling in a coolship, a metal vessel that is open to the environmental air. However, in-depth studies highlighted the role of the wooden barrels, in which the fermentation and maturation of these lambic beer production processes take place, for the inoculation of the wort. The present study took a deeper look into all possible inoculation sources of the lambic beer wort by sampling the raw materials, environmental air, and brewing surfaces and equipment, and performing two lambic beer production processes carried out with two traditional wheat landraces (Limburgse Rode and Witte van Vlaanderen). All samples were analyzed in both a culture-dependent and culture-independent way, and with respect to their metabolites. The results revealed that the wooden barrels contributed the most to the inoculation of the lambic beer wort, since they were responsible for the subsequent growth and activity of Saccharomyces and Brettanomyces yeast, lactic acid bacteria and acetic acid bacteria species, which all played their roles during lambic beer production processes. In addition, the house microbiota was identified as an inoculation source, mainly responsible for the contribution of microorganisms that play a role during the initial fermentation phase of lambic beer production processes.

Abstract Lac.to.ba.cil.la'ce.ae. N.L. masc. n. Lactobacillus , type genus of the family; L. fem. pl. n. suff. ‐aceae , ending to denote a family; N.L. fem. pl. n. Lactobacillaceae , the Lactobacillus family. Bacillota / “Bacilli” / Lactobacillales / Lactobacillaceae The emended family Lactobacillaceae was proposed based on phylogenomic analyses and encompassed all genera previously included in the families Lactobacillaceae Winslow et al. 1917 AL and Leuconostocaceae Schleifer 2010, VL132. This family currently comprises >420 non‐synonymous species classified into 36 genera and one Candidatus genus (July 2025). Cells are Gram‐stain‐positive, nonspore‐forming, facultatively or strictly anaerobic bacteria. Cells are coccoid or rod‐shaped may form chains, pairs, or in the case of Pediococcus, tetrads. The main product of the fermentative metabolism is lactate, and other products may be acetate, ethanol, CO 2 , formate, or succinate. Members of the Lactobacillaceae family have complex nutritional requirements for amino acids, peptides, nucleic acid derivatives, vitamins, salts, fatty acids or fatty acid esters, and fermentable carbohydrates. Lactobacillaceae is the only family in the order Lactobacillales that includes homofermentative and heterofermentative microorganisms. DNA G + C content (mol%) : 30–62 (genome sequences). Type genus : Lactobacillus Beijerinck 1901 AL emend. Haakensen et al. 2009, Cai et al. 2012, Zheng et al. 2020.

The use of beneficial microbes to reduce plant stress upon fungal pathogens is a promising strategy for sustainable crop protection. In this study, we evaluated five novel Burkholderia strains for their capacity to mitigate Botrytis cinerea infection in tomato (Solanum lycopersicum L.) and Fusarium graminearum infection in wheat (Triticum aestivum L.). All strains significantly suppressed B. cinerea infection symptoms in tomato plants, hallmarked by a smaller decrease in photosynthetic activity and chlorophyll content compared to infected control. Using a GFP-tagged F. graminearum strain, we show that the strains reduced fungal biomass accumulation in F. graminearum-infected wheat leaves and mitigated chlorosis. Phytotoxicity assessments revealed no adverse effects in tomato for any strain, while two strains induced a mild reduction in chlorophyll fluorescence in wheat, suggesting potential host-specific phytotoxicity. Whole-genome sequencing of the Burkholderia strains revealed a rich repertoire of biosynthetic gene clusters (BGCs) with conserved replicon positioning. While all strains shared BGCs for known bioactive metabolites such as occidiofungin A, ornibactin, and pyochelin, variation in other clusters did not directly correlate with the phenotypic effects observed. Our results highlight the strong biocontrol potential Burkholderia strains in two economically important crops against two globally important fungal pathounder controlled conditions. Finally, a microbial phylogenomic analysis revealed that the five strains belong to two new and previously uncharacterized species within the Burkholderia cepacia complex, for which the names Burkholderia mycopellens sp. nov. and Burkholderia crassaminum sp. nov. are proposed. These strains hold promise as next-generation biocontrol agents for enhancing crop health and managing fungal diseases sustainably.

Fresh-cut vegetables house a diverse microbial community, influenced by different pre- and post-harvest factors. The present study examined the temporal variability and the impact of a post-harvest washing process on the abundant and dominant microbial taxa of the indigenous microbiota of Equilibrium Modified-Atmosphere Packaged (EMAP) fresh-cut Romaine lettuce throughout the shelf-life. Whole heads of Romaine lettuce (Murcia, Spain) were sampled in triplicate in January, June and October 2022. Manually cut leaves were washed in potable water and a 200 mg/L chlorine solution, respectively, and stored in EMAP conditions (3 % O-2, 10 % CO2) at 7 degrees C for seven days. Quantitative data were obtained through plating on selective and non-selective culture media, followed by bacterial isolation and taxonomic identification using MALDI-TOF MS and 16S rRNA gene sequencing. The psychrotrophic aerobic plate count of unwashed lettuce ranged from 5.1 to 6.3 log CFU/g, reaching up to 7.5 log CFU/g by the end of the shelf-life. The 808 bacterial isolates were dominated by species from the phylum Pseudomonadota (90 %), with Pseudomonas and Pantoea as the core genera. Temporal variability on the microbiota was demonstrated, with a 1-log lower microbial load in June and the community composition varying across all sampling periods. The chlorine wash induced a reduction in total count up to 2.2 log CFU/g, but had no significant quantitative, nor qualitative impact on the microbial community throughout the shelf-life. These findings suggest that pre-harvest environmental conditions play a more crucial role than post-harvest washing in determining the variability of the Romaine lettuce microbiota.

The vaginal microbiome (VMB) plays a crucial role in women’s reproductive health, yet its composition and variability among different ethnic groups in Armenia have not been previously investigated using metagenomic approaches. This study aimed to characterize and compare the VMB of women from three ethnic groups—Armenians, Yezidis, and Molokans—using classical microbiological methods, MALDI-TOF MS, and 16S rRNA amplicon-based metagenomic sequencing. The analyses revealed a high level of consistency between culture-based and metagenomic data. Among the studied groups, Armenian women exhibited the greatest microbial diversity at both species and phylum levels. In women with candidiasis, a marked reduction in lactobacilli abundance and an increased presence of pathogenic species were observed. Pregnant women generally displayed a balanced, lactobacilli-dominated microbiome, although opportunistic pathogens were occasionally detected, likely due to immune modulation during pregnancy. Regional and ethnic differences were most pronounced at the phylum level. These findings represent the first comprehensive metagenomic assessment of the VMB in Armenia, providing important insights into microbial diversity across ethnic groups and its implications for women’s reproductive health.

Actinotignum schaalii is a human commensal bacterium that is well-known for causing urinary tract infections, predominantly in the elderly. However, reliable identification of A. schaalii in routine clinical microbiology has been challenging, mainly due to its fastidious growth and identification issues. The genus Actinotignum currently comprises three species with validly published names, i.e., A. schaalii, A. sanguinis, and A. urinale. The former two are difficult to distinguish by routine clinical microbiology laboratory protocols, including phenotypic profiling and matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS). A total of 137 new Actinotignum isolates were collected from urine and various other human clinical samples between 2010 and 2023 and were identified by MALDI-TOF MS as A. schaalii/A. sanguinis. Whole-genome sequence-based analyses identified none of these isolates as A. schaalii or A. sanguinis and demonstrated that the strains were clustered into seven genomic groups. These seven genomic groups represented seven novel Actinotignum species, which we propose to be classified as A. lotii sp. nov. (n = 82 isolates), A. saccati sp. nov. (n = 38), A. inguinis sp. nov. (n = 8), A. vesiculae sp. nov. (n = 3), A. stranguriae sp. nov. (n = 3), A. cystesis sp. nov. (n = 2), and A. urematis sp. nov. (n = 1). Therefore, MALDI-TOF MS analysis allows reliable identification only to the genus level for Actinotignum species, whereas commercial identification libraries need revision and extension to improve their diagnostic capacities.IMPORTANCEActinotignum schaalii is a human commensal bacterium that causes urinary tract infections, as well as invasive infections, predominantly in the elderly. Yet, isolation and reliable identification of A. schaalii in routine clinical microbiology is challenging. Today, matrix-assisted laser desorption/ionization mass spectrometry (MALDI-TOF MS) has been recommended for the identification of Actinotignum bacteria but fails to discriminate between A. schaalii and A. sanguinis. The present study used whole-genome sequence analyses and demonstrated that, among 137 new human clinical Actinotignum isolates tentatively identified as A. schaalii/A. sanguinis through MALDI-TOF MS, none belonged to A. schaalii or A. sanguinis; rather, they represent seven novel Actinotignum species. The commercial MALDI-TOF MS identification databases need marked improvement if correct species-level identifications will be reliable. Currently, whole-genome sequence-based identification may need to be used to provide correct species identifications and to assess the genetic and clinical differences between established and newly defined Actinotignum species.

Research background. Fermented soymilk has emerged as a potential functional food due to its nutritional and health-promoting properties. Enhancing its functionality by enriching it with gamma-aminobutyric acid (GABA), a neurotransmitter with various health benefits, is an area of active research. This study aims to develop a novel GABA-enriched fermented soymilk using a newly isolated Lactiplantibacillus plantarum strain with both probiotic and GABA-producing capabilities. Experimental approach. Five L. plantarum strains isolated from Vietnamese soybean whey were screened for GABA production and probiotic characteristics. Strain W12, which exhibited superior performance, was selected for optimisation. Response surface methodology with a central composite design was used to optimise monosodium glutamate (MSG) and sucrose amounts for maximal GABA yield. A time-course study was then conducted to monitor bacterial growth, pH changes, organic acid production, and GABA accumulation during fermentation under the optimised conditions. Results and conclusions. L. plantarum W12 demonstrated exceptional probiotic traits: 97.1 % survival at pH=2.5, 96.5 % survival in bile salts/pancreatin, 97.3 % pepsin tolerance, 96.7 % auto-aggregation and 85.4 % hydrophobicity in chloroform. Initial GABA production reached (9.1±0.4) mM. Response surface optimisation predicted a maximum GABA concentration of 34.2 mM at 1.564 mg/mL MSG and 10.93 % sucrose, which was experimentally validated at (34.5±1.0) mM (15 h, 45 °C). Lactic acid was the predominant organic acid produced ((182.4±8.0) mg/g at 18 h), with viable cell counts exceeding 7.9 log CFU/mL, meeting probiotic thresholds. Novelty and scientific contribution. This study presents the first comprehensive characterisation of L. plantarum W12, a strain combining exceptional GABA biosynthesis with robust probiotic properties. The systematic response surface methodology (RSM) optimisation framework and detailed metabolic profiling provide reproducible protocols for developing multifunctional fermented soy beverages with applications in neurological and gastrointestinal health promotion.

The taxonomy of the genus Arcobacter has been subject to substantive turmoil in recent years following a proposal to subdivide the genus into six genera. This proposal has been challenged by a number of multidisciplinary studies employing phenotypic, genomic, and phylogenetic analyses. Following several discussions among members of the International Committee on Systematics of Prokaryotes (ICSP) subcommittee on the taxonomy of Campylobacter and related bacteria, this group now unanimously recommends the use of the genus term Arcobacter to refer to these species.

For the last decade, matrix-assisted laser desorption/ionization time-of-flight mass spectrometry (MALDI-TOF MS) has been the reference method for species identification in clinical microbiology. Hampered by a historical lack of open data, machine learning research towards models specifically adapted to MALDI-TOF MS remains in its infancy. Given the growing complexity of available datasets (such as large-scale antimicrobial resistance prediction), a need for models that (1) are specifically designed for MALDI-TOF MS data, and (2) have high representational capacity, presents itself.Here, we introduce Maldi Transformer, an adaptation of the state-of-the-art transformer architecture to the MALDI-TOF mass spectral domain. We propose the first self-supervised pre-training technique specifically designed for mass spectra. The technique is based on shuffling peaks across spectra, and pre-training the transformer as a peak discriminator. Extensive benchmarks confirm the efficacy of this novel design. The final result is a model exhibiting state-of-the-art (or competitive) performance on downstream prediction tasks. In addition, we show that Maldi Transformer’s identification of noisy spectra may be leveraged towards higher predictive performance.All code supporting this study is distributed on PyPI and is packaged under: https://github.com/gdewael/maldi-nn.

A survey of 75 commercially available, unpasteurised, spontaneously fermented vegetables was conducted, focusing on the culturable microbiota. Foodborne pathogens (Listeria monocytogenes, Salmonella sp.), indicator bacteria (generic E. coli, Enterobacteriaceae), and lactic acid bacteria (LAB) were analysed both at the day of purchase and at the end of shelf life. Selected samples were further examined for biogenic amines, and dominant culturable LAB were isolated and identified. Challenge tests evaluated the growth potential of L. monocytogenes during white cabbage fermentation. Results showed considerable variation in pH (3.1-4.3) and LAB counts (<1.0-8.8 log CFU/g), reflecting the spontaneous nature of the vegetable fermentations. No Salmonella sp. (detection in 25 g), L. monocytogenes (LOD 10 CFU/g), or E. coli (LOD 10 CFU/g) were detected, indicating no immediate food safety or hygiene concerns, although Enterobacteriaceae were present (LOD 10 CFU/g) in six samples, suggesting an inadequate (spontaneous) fermentation process to ensure die-off of enteric pathogens. LAB isolates were dominated by Pediococcus parvulus, Lactiplantibacillus plantarum, Levilactobacillus brevis, and Lentilactobacillus buchneri. High biogenic amine levels highlight an increased risk for adverse health effects related to spontaneous vegetable fermentations. Challenge tests confirmed that rapid acidification to pH <4.4, followed by a minimum 14-day holding period under optimal fermentation conditions, effectively limits pathogen outgrowth and/or survival. Good fermentation practices, together with a sensory evaluation are essential to ensure the microbiological safety, quality, and consistency of spontaneous vegetable fermentations.

The genus Burkholderia belongs to the class Betaproteobacteria within the phylum of the Pseudomonadota. Although recent taxonomic revisions have reduced the number of recognized Burkholderia species from several hundred to a few dozen, this genus remains exceptionally diverse and functionally versatile. This versatility is supported by large genomes up to 10 Mb in size typically organized into three replicons. Burkholderia sensu stricto species have been isolated from a wide range of sources, but accumulating evidence indicates that soil - where they associate with various plants, fungi, and protists - is their natural habitat. The genus includes two class 3 pathogens, Burkholderia mallei and Burkholderia pseudomallei, which cause glanders in horses and melioidosis in humans, respectively. In addition, an increasing number of Burkholderia species have been identified as opportunistic human pathogens. These bacteria show a particular affinity for the respiratory tract, with life-threatening lung infections caused by Burkholderia cepacia complex bacteria occurring primarily in individuals with cystic fibrosis or those requiring mechanical ventilation. Burkholderia species have been described as both friends and foes of humans. While they are perhaps best known for their roles as pathogens of humans, animals, and plants, numerous studies have also highlighted their biotechnological potential, in particular, their abilities to promote plant growth, control agricultural pests, and contribute to bioremediation efforts. Today, an exceptionally large number of Burkholderia whole-genome sequences are available. However, these data have generally failed to clearly distinguish between beneficial and harmful strains, or between those considered safe or unsafe for biotechnological applications. Nonetheless, genomic analyses combined with in planta experiments - now underpin a growing number of studies aimed at harnessing this genetic potential as a source of novel antibiotics and other bioactive secondary metabolites.

We used matrix-assisted laser desorption/ionization time-of-flight mass spectrometry and whole-genome sequence analyses to identify 90 Bartonella isolates from honey bee gut samples in Belgium. While the identification of 62 isolates as Bartonella apihabitans and three as Bartonella choladocola was straightforward, the identification of 25 Bartonella apis-like isolates was challenging. A taxonomic and functional analysis of four B. apis-like genomes and of publicly available B. apis genomes demonstrated that neither OrthoANIu and digital DNA-DNA hybridization analyses, nor functional annotation supported a clear separation of B. apis and B. apis-like genomes. Different phylogenomic analyses showed that B. apis and B. apis-like strains formed a monophyletic clade with an inconsistent internal structure. We therefore considered the remaining 25 isolates identified as B. apis. We subsequently re-addressed an earlier phylogenetic and functional divergence between three major clades of Bartonella species which differed not only in phylogenomic position and ecology, but also in genome size and genomic percentage G + C content, and in many metabolic capabilities. We propose to reclassify the single species of the Bartonella tamiae clade into the novel genus Attibartonella gen. nov., with Attibartonella tamiae comb. nov. as the type species. Similarly, we propose to reclassify species of the honey bee-associated Bartonella clade into the novel genus Ditibartonella gen. nov., with Ditibartonella apis comb. nov. as the type species. The phylogenomic analyses of publicly available genome and metagenome sequences revealed additional Ditibartonella species in honey bee samples, highlighted an evolutionary adaptation of Ditibartonella bacteria to bee hosts and suggested shared transmission routes.

A Gram-stain-negative coccobacillus, LMG 32992 T , was isolated from water that had collected in a tyre in Pingtung, Donggang Township, Taiwan. Upon preliminary 16S rRNA gene sequence analysis, it was most closely related to members of the genus Ralstonia (16S rRNA gene sequence similarities of 96.7–97.5%). The present study aimed to elucidate its taxonomic position and to propose a formal classification. To this end, the complete genome sequence was determined, and taxonomic, phylogenomic, metabolic and physiological analyses were performed. Comparative genomic analyses demonstrated that strain LMG 32992 T and another unclassified strain, Burkholderiaceae bacterium PBA, which was isolated earlier from textile wastewater in Malaysia, represented a single novel species within a novel genus of the family Burkholderiaceae . The G+C content of the LMG 32992 T genomic DNA was 63.77 mol%. Genomic analyses and growth tests demonstrated that LMG 32992 T had an asaccharolytic metabolism but that it was well-equipped to synthetize, if necessary autotrophically, and transform all required carbohydrates and that it used the Krebs and related cycles to generate reductive power for a heterotrophic energy metabolism. We propose the name Imbroritus primus gen. nov., sp. nov. with strain LMG 32992 T (=CIP 112179 T =BCRC 81361 T =A30B1 T ) as the type strain, for this novel taxon.

Fifteen isolates from gut samples of the invasive solitary bee Megachile sculpturalis remained unidentified after matrix-assisted laser desorption/ionization time-of-flight mass spectrometry analysis. Phylogenomic and overall genome relatedness indices analyses of five representative isolates demonstrated that 14 isolates represented a novel Carnimonas species for which we propose the name Carnimonas bestiolae sp. nov., with LMG 33810T as the type strain. A single isolate represented a distinct lineage within the Halomonadaceae family with the genera Carnimonas and Halotalea as nearest neighbor taxa and is classified as Cernens ardua gen. nov., sp. nov., with LMG 33818T as the type strain. Comparative genomic analysis and physiological characterization revealed functional differences between C. bestiolae, Ce. ardua and established Carnimonas and Halotalea species in central metabolism, fatty acid metabolism, and osmoregulatory pathways, the latter being consistent with adaptation to a saline environment. A reanalysis of previously published insect microbiome studies in which Carnimonas was reported, revealed 16S rRNA amplicon sequence variants with >99 % identity to C. bestiolae LMG 33810T 16S rRNA in gut samples of Megachile sculpturalis, Anthidium florentinum, and the psyllid Leptynoptera sulfurea. In contrast, 16S rRNA amplicon sequence variants corresponding to Ce. ardua were rarely detected and only at low relative abundances, suggesting a transient association with the insect gut.

Standard antibiotic susceptibility tests (ASTs) often fail to accurately predict treatment outcomes because they do not account for biofilm-specific mechanisms of reduced susceptibility. In the present study, we explored alternative approaches to predict tobramycin susceptibility of Pseudomonas aeruginosa biofilms that were experimentally evolved in physiologically relevant conditions. To this end, we used four analytical methods – whole-genome sequencing (WGS), matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS), isothermal microcalorimetry (IMC) and multi-excitation Raman spectroscopy (MX-Raman). Machine learning models were trained on data outputs from these methods to predict tobramycin susceptibility of our evolved strains and subsequently validated with a collection of clinical isolates. For minimal inhibitory concentration (MIC) predictions of the evolved strains, the highest accuracy ±1 was achieved with MALDI-TOF MS (97.83 %), while for biofilm prevention concentration (BPC) predictions, Raman spectroscopy performed best with an accuracy ±1 of 80.43 %. Overall, all analytical methods demonstrated comparable predictive performance, showing their potential for improving biofilm AST.

As defined by law, at least 30 % of the total amount of starch- or carbohydrate-containing raw materials used for the production of Belgian lambic beers should be unmalted wheat, since this is important for the longevity of the lambic beer production process. Whereas lambic beer brewers in the 19th and 20th century depended on local wheat landraces, supplied by the local farmers, they nowadays use modern wheat varieties that are mainly cultivated for their baking quality and available on the global market. However, it could be assumed that lambic beers made with traditional wheat varieties were characterized by different flavor profiles. To know the impact of those wheat varieties on the growing microbiota and metabolites produced throughout a lambic beer production process, three different productions were assessed as a function of fermentation and maturation time. Two of them were performed with a traditional wheat landrace, namely, Witte van Vlaanderen and Limburgse Rode, and one control was performed with current standard wheat. A multiphasic analysis approach revealed that the impact of the wheat variety on the microbiology and metabolome was limited, although some metabolic differences, such as the concentrations of volatile phenolic compounds and esters, could be linked with the use of the Limburgse Rode wheat landrace. However, the most impacting factor remained the spontaneous origin of the prevailing microorganisms and the environmental conditions, such as temperature, pH, and dissolved oxygen concentration, that are encountered during lambic beer production. Finally, this work allowed to discover a new species, namely, Acetobacter zythi.

Antibiotic susceptibility tests (ASTs) often fail to predict treatment outcomes because they do not account for biofilm-specific tolerance mechanisms. In the present study, we explored alternative approaches to predict tobramycin susceptibility of Pseudomonas aeruginosa biofilms that were experimentally evolved in physiologically relevant conditions. To this end, we used four analytical methods - whole-genome sequencing (WGS), matrix-assisted laser desorption/ionization-time of flight mass spectrometry (MALDI-TOF MS), isothermal microcalorimetry (IMC) and multi-excitation Raman spectroscopy (MX-Raman). Machine learning models were trained on data outputs from these methods to predict tobramycin susceptibility of our evolved strains and validated with a collection of clinical isolates. For minimal inhibitory concentration (MIC) predictions of the evolved strains, the highest accuracy was achieved with MALDI-TOF MS (97.83%), while for biofilm prevention concentration (BPC) predictions, Raman spectroscopy performed best with an accuracy of 80.43%. Overall, all analytical methods demonstrated comparable predictive performance, showing their potential for improving biofilm AST.