Exploring the gut DNA virome in fecal immunochemical test stool samples reveals associations with lifestyle in a large population-based study.

Stool samples for fecal immunochemical tests (FIT) are collected in large numbers worldwide as part of colorectal cancer screening programs. Employing FIT samples from 1034 CRCbiome participants, recruited from a Norwegian colorectal cancer screening study, we identify, annotate and characterize more than 18000 DNA viruses, using shotgun metagenome sequencing. Only six percent of them are assigned to a known taxonomic family, with Microviridae being the most prevalent viral family. Linking individual profiles to comprehensive lifestyle and demographic data shows 17/25 of the variables to be associated with the gut virome. Physical activity, smoking, and dietary fiber consumption exhibit strong and consistent associations with both diversity and relative abundance of individual viruses, as well as with enrichment for auxiliary metabolic genes. We demonstrate the suitability of FIT samples for virome analysis, opening an opportunity for large-scale studies of this enigmatic part of the gut microbiome. The diverse viral populations and their connections to the individual lifestyle uncovered herein paves the way for further exploration of the role of the gut virome in health and disease.

Clinical Diagnostics of Bacterial Infections and Their Resistance to Antibiotics-Current State and Whole Genome Sequencing Implementation Perspectives.

Antimicrobial resistance (AMR), defined as the ability of microorganisms to withstand antimicrobial treatment, is responsible for millions of deaths annually. The rapid spread of AMR across continents warrants systematic changes in healthcare routines and protocols. One of the fundamental issues with AMR spread is the lack of rapid diagnostic tools for pathogen identification and AMR detection. Resistance profile identification often depends on pathogen culturing and thus may last up to several days. This contributes to the misuse of antibiotics for viral infection, the use of inappropriate antibiotics, the overuse of broad-spectrum antibiotics, or delayed infection treatment. Current DNA sequencing technologies offer the potential to develop rapid infection and AMR diagnostic tools that can provide information in a few hours rather than days. However, these techniques commonly require advanced bioinformatics knowledge and, at present, are not suited for routine lab use. In this review, we give an overview of the AMR burden on healthcare, describe current pathogen identification and AMR screening methods, and provide perspectives on how DNA sequencing may be used for rapid diagnostics. Additionally, we discuss the common steps used for DNA data analysis, currently available pipelines, and tools for analysis. Direct, culture-independent sequencing has the potential to complement current culture-based methods in routine clinical settings. However, there is a need for a minimum set of standards in terms of evaluating the results generated. Additionally, we discuss the use of machine learning algorithms regarding pathogen phenotype detection (resistance/susceptibility to an antibiotic).

Ultrafast and Cost-Effective Pathogen Identification and Resistance Gene Detection in a Clinical Setting Using Nanopore Flongle Sequencing.

Rapid bacterial identification and antimicrobial resistance gene (ARG) detection are crucial for fast optimization of antibiotic treatment, especially for septic patients where each hour of delayed antibiotic prescription might have lethal consequences. This work investigates whether the Oxford Nanopore Technology's (ONT) Flongle sequencing platform is suitable for real-time sequencing directly from blood cultures to identify bacteria and detect resistance-encoding genes. For the analysis, we used pure bacterial cultures of four clinical isolates of Escherichia coli and Klebsiella pneumoniae and two blood samples spiked with either E. coli or K. pneumoniae that had been cultured overnight. We sequenced both the whole genome and plasmids isolated from these bacteria using two different sequencing kits. Generally, Flongle data allow rapid bacterial ID and resistome detection based on the first 1,000-3,000 generated sequences (10 min to 3 h from the sequencing start), albeit ARG variant identification did not always correspond to ONT MinION and Illumina sequencing-based data. Flongle data are sufficient for 99.9% genome coverage within at most 20,000 (clinical isolates) or 50,000 (positive blood cultures) sequences generated. The SQK-LSK110 Ligation kit resulted in higher genome coverage and more accurate bacterial identification than the SQK-RBK004 Rapid Barcode kit.

Hybrid Assembly Provides Improved Resolution of Plasmids, Antimicrobial Resistance Genes, and Virulence Factors in Escherichia coli and Klebsiella pneumoniae Clinical Isolates.

Emerging new sequencing technologies have provided researchers with a unique opportunity to study factors related to microbial pathogenicity, such as antimicrobial resistance (AMR) genes and virulence factors. However, the use of whole-genome sequence (WGS) data requires good knowledge of the bioinformatics involved, as well as the necessary techniques. In this study, a total of nine Escherichia coli and Klebsiella pneumoniae isolates from Norwegian clinical samples were sequenced using both MinION and Illumina platforms. Three out of nine samples were sequenced directly from blood culture, and one sample was sequenced from a mixed-blood culture. For genome assembly, several long-read, (Canu, Flye, Unicycler, and Miniasm), short-read (ABySS, Unicycler and SPAdes) and hybrid assemblers (Unicycler, hybridSPAdes, and MaSurCa) were tested. Assembled genomes from the best-performing assemblers (according to quality checks using QUAST and BUSCO) were subjected to downstream analyses. Flye and Unicycler assemblers performed best for the assembly of long and short reads, respectively. For hybrid assembly, Unicycler was the top-performing assembler and produced more circularized and complete genome assemblies. Hybrid assembled genomes performed substantially better in downstream analyses to predict putative plasmids, AMR genes and ß-lactamase gene variants, compared to MinION and Illumina assemblies. Thus, hybrid assembly has the potential to reveal factors related to microbial pathogenicity in clinical and mixed samples.

Fighting Antibiotic Resistance in Hospital-Acquired Infections: Current State and Emerging Technologies in Disease Prevention, Diagnostics and Therapy.

Rising antibiotic resistance is a global threat that is projected to cause more deaths than all cancers combined by 2050. In this review, we set to summarize the current state of antibiotic resistance, and to give an overview of the emerging technologies aimed to escape the pre-antibiotic era recurrence. We conducted a comprehensive literature survey of >150 original research and review articles indexed in the Web of Science using "antimicrobial resistance," "diagnostics," "therapeutics," "disinfection," "nosocomial infections," "ESKAPE pathogens" as key words. We discuss the impact of nosocomial infections on the spread of multi-drug resistant bacteria, give an overview over existing and developing strategies for faster diagnostics of infectious diseases, review current and novel approaches in therapy of infectious diseases, and finally discuss strategies for hospital disinfection to prevent MDR bacteria spread.

AMR-Diag: Neural network based genotype-to-phenotype prediction of resistance towards ß-lactams in Escherichia coli and Klebsiella pneumoniae.

Antibiotic resistance poses a major threat to public health. More effective ways of the antibiotic prescription are needed to delay the spread of antibiotic resistance. Employment of sequencing technologies coupled with the use of trained neural network algorithms for genotype-to-phenotype prediction will reduce the time needed for antibiotic susceptibility profile identification from days to hours. In this work, we have sequenced and phenotypically characterized 171 clinical isolates of Escherichia coli and Klebsiella pneumoniae from Norway and India. Based on the data, we have created neural networks to predict susceptibility for ampicillin, 3rd generation cephalosporins and carbapenems. All networks were trained on unassembled data, enabling prediction within minutes after the sequencing information becomes available. Moreover, they can be used both on Illumina and MinION generated data and do not require high genome coverage for phenotype prediction. We cross-checked our networks with previously published algorithms for genotype-to-phenotype prediction and their corresponding datasets. Besides, we also created an ensemble of networks trained on different datasets, which improved the cross-dataset prediction compared to a single network. Additionally, we have used data from direct sequencing of spiked blood cultures and found that AMR-Diag networks, coupled with MinION sequencing, can predict bacterial species, resistome, and phenotype as fast as 1-8 h from the sequencing start. To our knowledge, this is the first study for genotype-to-phenotype prediction: (1) employing a neural network method; (2) using data from more than one sequencing platform; and (3) utilizing sequence data from spiked blood cultures.

Plasmid Identification and Plasmid-Mediated Antimicrobial Gene Detection in Norwegian Isolates.

Norway is known for being one of the countries with the lowest levels of antimicrobial resistance (AMR). AMR, through acquired genes located on transposons or conjugative plasmids, is the horizontal transmission of genes required for a given bacteria to withstand antibiotics. In this work, bioinformatic analysis of whole-genome sequences and hybrid assembled data from Escherichia coli, and Klebsiella pneumoniae isolates from Norwegian patients was performed. For detection of putative plasmids in isolates, the plasmid assembly mode in SPAdes was used, followed by annotation of resulting contigs using PlasmidFinder and two curated plasmid databases (Brooks and PLSDB). Furthermore, ResFinder and Comprehensive Antibiotic Resistance Database (CARD) were used for the identification of antibiotic resistance genes (ARGs). The IncFIB plasmid was detected as the most prevalent plasmid in both E. coli, and K. pneumoniae isolates. Furthermore, ARGs such as aph(3″)-Ib, aph(6)-Id, sul1, sul2, tet(D), and qnrS1 were identified as the most abundant plasmid-mediated ARGs in Norwegian E. coli and K. pneumoniae isolates, respectively. Using hybrid assembly, we were able to locate plasmids and predict ARGs more confidently. In conclusion, plasmid identification and ARG detection using whole-genome sequencing data are heavily dependent on the database of choice; therefore, it is best to use several tools and/or hybrid assembly for obtaining reliable identification results.

Early Gut Fungal and Bacterial Microbiota and Childhood Growth.

Introduction: Childhood growth is a sensitive marker of health. Animal studies show increased height and weight velocity in the presence of fungal as well as antibiotic supplement in feed. Human studies on early gut microbiota and anthropometrics have mainly focused on bacteria only and overweight, with diverging results. We thus aimed to investigate the associations between childhood growth [height and body mass index (BMI)] and early fungal and bacterial gut microbiota. Methods: In a population-based cohort, a subset of 278 pregnant mothers was randomized to drink milk with or without probiotic bacteria during and after pregnancy. We obtained fecal samples in offspring at four time points between 0 and 2 years and anthropometric measurements 0 and 9 years. By quantitative PCR and 16S/ITS rRNA gene sequencing, children's gut microbiota abundance and diversity were analyzed against height standard deviation score (SDS) and BMI-SDS and presented as effect estimate (ß) of linear mixed models. Results: From 278 included children (149 girls), 1,015 fecal samples were collected. Maternal probiotic administration did not affect childhood growth, and the groups were pooled. Fungal abundance at 2 years was positively associated with height-SDS at 2-9 years (ß = 0.11 height-SDS; 95% CI, 0.00, 0.22) but not with BMI-SDS. Also, higher fungal abundance at 1 year was associated with a lower BMI-SDS at 0-1 year (ß = -0.09 BMI-SDS; 95% CI, -0.18, -0.00), and both bacterial abundance and bacterial alpha diversity at 1 year were associated with lower BMI-SDS at 0-1 year (ß = -0.13 BMI-SDS; 95% CI, -0.22, -0.04; and ß = -0.19 BMI-SDS; 95% CI, -0.39, -0.00, respectively). Conclusions: In this prospective cohort following 0-9-year-old children, we observed that higher gut fungal abundances at 2 years were associated with taller children between 2 and 9 years. Also, higher gut fungal and bacterial abundances and higher gut bacterial diversity at 1 year were associated with lower BMI in the first year of life. The results may indicate interactions between early gut fungal microbiota and the human growth-regulating physiology, previously not reported. Clinical Trial Registration: Clinicaltrials.gov, NCT00159523.

Rapid identification of pathogens, antibiotic resistance genes and plasmids in blood cultures by nanopore sequencing.

Bloodstream infections (BSI) and sepsis are major causes of morbidity and mortality worldwide. Blood culture-based diagnostics usually requires 1-2 days for identification of bacterial agent and an additional 2-3 days for phenotypic determination of antibiotic susceptibility pattern. With the escalating burden of antimicrobial resistance (AMR) rapid diagnostics becomes increasingly important to secure adequate antibiotic therapy. Real-time whole genome sequencing represents a genotypic diagnostic approach with the ability to rapidly identify pathogens and AMR-encoding genes. Here we have used nanopore sequencing of bacterial DNA extracted from positive blood cultures for identification of pathogens, detection of plasmids and AMR-encoding genes. To our knowledge, this is the first study to gather the above-mentioned information from nanopore sequencing and conduct a comprehensive analysis for diagnostic purposes in real-time. Identification of pathogens was possible after 10 minutes of sequencing and all predefined AMR-encoding genes and plasmids from monoculture experiments were detected within one hour using raw nanopore sequencing data. Furthermore, we demonstrate the correct identification of plasmids and blaCTX-M subtypes using de novo assembled nanopore contigs. Results from this study hold great promise for future applications in clinical microbiology and for health care surveillance purposes.

Culture dependent and independent analyses suggest a low level of sharing of endospore-forming species between mothers and their children.

Spore forming bacteria comprise a large part of the human gut microbiota. However, study of the endospores in gut microbiota is limited due to difficulties of culturing and numerous unknown germination factors. In this study we propose a new method for culture-independent characterization of endospores in stool samples. We have enriched DNA of spore-forming bacterial species from stool samples of 40 mother-child pairs from a previously described mother-child cohort. The samples were exposed to a two-step purification process comprising ethanol and ethidium monoazide (EMA) treatment to first kill vegetative cells and to subsequently eliminate their DNA from the samples. The composition of the ethanol-EMA resistant DNA was characterized by 16S rRNA marker gene sequencing. Operational taxonomic units (OTUs) belonging to the Clostridia class (OTU1: Romboutsia, OTU5: Peptostreptococcaceae and OTU14: Clostridium senso stricto) and one belonging to the Bacillus class (OTU20: Turicibacter) were significantly more abundant in the samples from mothers and children after ethanol-EMA treatment than in those treated with ethanol only. No correlation was observed between ethanol-EMA resistant OTUs detected in children and in their mothers, which indicates that a low level of spore-forming species are shared between mothers and their children. Anaerobic ethanol-resistant bacteria were isolated from all mothers and all children over 1 year of age. Generally, in 70% of the ethanol-treated samples used for anaerobic culturing, 16S rRNA gene sequences of bacterial isolates corresponded to OTUs detected in these samples after EMA treatment. We report a new DNA-based method for the characterization of endospores in gut microbiota. Our method has high degree of correspondence to the culture-based method, although it requires further optimization. Our results also indicate a high turnover of endospores in the gut during the first two years of life, perhaps with a high environmental impact.

Exploring the Brine Microbiota of a Traditional Norwegian Fermented Fish Product (Rakfisk) from Six Different Producers during Two Consecutive Seasonal Productions.

The purpose of this study was to explore the microbiota of Norwegian fermented fish (rakfisk), a traditional product popular in the Norwegian market. Brine samples, collected from six producers during two subsequent years, were used. The producers applied different salt concentrations (between 3.8% and 7.2% NaCl), ripening temperatures (between 3.5 and 7.5 °C), fish species (trout or char), and fish upbringing (wild trout, on-shore farmed trout or char, and off-shore farmed char). The microbiota in the brine during the ripening process was mainly characterized by DNA-based, culture-independent methods. In total, 1710 samples were processed and of these 1342 were used for the final analysis. The microbiota was dominated by Gammaproteobacteria and Bacilli with the largest variance between samples associated with the genera Psychrobacter and Lactobacillus. The variance in the material was mainly determined by the origin of the samples, i.e., the different producers. The microbiota from the individual producers was to a large extent reproducible from one year to the next and appeared to be determined by the relatively small differences in the salinity and the ripening temperature. This is the first study exploring the microbiota in rakfisk brine and it provides insights into environmental factors affecting the rakfisk ecosystems.

Low Maternal Microbiota Sharing across Gut, Breast Milk and Vagina, as Revealed by 16S rRNA Gene and Reduced Metagenomic Sequencing.

The maternal microbiota plays an important role in infant gut colonization. In this work we have investigated which bacterial species are shared across the breast milk, vaginal and stool microbiotas of 109 women shortly before and after giving birth using 16S rRNA gene sequencing and a novel reduced metagenomic sequencing (RMS) approach in a subgroup of 16 women. All the species predicted by the 16S rRNA gene sequencing were also detected by RMS analysis and there was good correspondence between their relative abundances estimated by both approaches. Both approaches also demonstrate a low level of maternal microbiota sharing across the population and RMS analysis identified only two species common to most women and in all sample types (Bifidobacterium longum and Enterococcus faecalis). Breast milk was the only sample type that had significantly higher intra- than inter- individual similarity towards both vaginal and stool samples. We also searched our RMS dataset against an in silico generated reference database derived from bacterial isolates in the Human Microbiome Project. The use of this reference-based search enabled further separation of Bifidobacterium longum into Bifidobacterium longum ssp. longum and Bifidobacterium longum ssp. infantis. We also detected the Lactobacillus rhamnosus GG strain, which was used as a probiotic supplement by some women, demonstrating the potential of RMS approach for deeper taxonomic delineation and estimation.

Comparison of reduced metagenome and 16S rRNA gene sequencing for determination of genetic diversity and mother-child overlap of the gut associated microbiota.

Use of the 16S rRNA gene in microbiota studies is limited by the lack of taxonomic and functional resolution. High resolution analyses are particularly important for understanding transmission and persistence of bacteria. The aim of our work was therefore to compare a novel reduced metagenome sequencing (RMS) approach with 16S rRNA gene sequencing to determine both the metagenome genetic diversity and the mother-to-child sharing of the microbiota in a cohort of 17 mother-child pairs. We found that although both approaches gave comparable results with respect to sample separation and taxonomy, RMS gave higher resolution and the potential for genomic-/functional assignment. Using RMS we estimated that the metagenome size increased from about 60 Mbp for 4-day-old children to about 225 Mbp for mothers. The 4-day-old children shared 7% of the metagenome sequences with the mothers, while the metagenome sequence sharing was >30% among the mothers. We found 15 genomes shared across >50% of the mothers, of which 10 belonged to Clostridia. Only Bacteroides showed a direct mother-child association, with B. vulgatus being abundant in both 4-day-old children and mothers. For the functional assignments, we identified a significant association between antibiotic usage during labor, and quantity of Fosfomycin resistance genes. In conclusion, our results show a higher functional and taxonomic resolution for RMS compared to 16S rRNA gene sequencing, where RMS enabled a detailed description of mother to child gut microbiota transmission - supporting a late recruitment of most gut bacteria and an effect of antibiotic treatment during labor on infant antibiotic resistance gene patterns.

Breastfeeding-associated microbiota in human milk following supplementation with Lactobacillus rhamnosus GG, Lactobacillus acidophilus La-5, and Bifidobacterium animalis ssp. lactis Bb-12.

Breastfeeding is one of the major factors affecting the early development of the infant gut microbiota, and weaning is associated with a shift in the gut microbiota toward a more adult composition. Through breastfeeding, infants receive bioactive components that shape their microbiota while also being exposed to the breast milk and breast surface microbial communities. Recent studies have suggested the possibility of an entero-mammary route of microbial transfer, opening the possibility of infant gut microbiota modulation through maternal probiotic supplementation. In this study, we have analyzed breast milk samples collected at 10 d and 3 mo postpartum from women participating in the Probiotics in the Prevention of Allergy among Children in Trondheim placebo controlled trial. Women who were randomized to the probiotic arm of the Probiotics in the Prevention of Allergy among Children in Trondheim trial received a fermented milk supplemented with Lactobacillus rhamnosus GG, Lactobacillus acidophilus La-5, and Bifidobacterium animalis ssp. lactis Bb-12, consuming this daily from 4 wk before their expected due date until 3 mo after birth. In total, 472 breast milk samples were assessed for the administered bacteria using quantitative real-time PCR and the microbiota transferred during breastfeeding was analyzed using 16S ribosomal RNA gene sequencing of 142 samples. We found that breastfeeding is unlikely to be a significant source of L. rhamnosus GG, L. acidophilus La-5, and B. animalis ssp. lactis Bb-12 for infants in the probiotic arm of the trial. Furthermore, maternal supplementation did not significantly affect the overall composition of the breast milk microbiota transferred during breastfeeding. We also present a descriptive analysis of this microbiota, which was largely dominated by Streptococcus and Staphylococcus genera at both 10 d and 3 mo postpartum. Samples collected at 3 mo postpartum had a statistically significant lower presence and relative abundance of the Staphylococcus genus. These samples also had a greater number of observed species and diversity, including more operational taxonomic units from the Rothia, Veillonella, Granulicatella, and Methylbacterium genera.

Diversity of vaginal microbiota increases by the time of labor onset.

Vaginal microbiota is an important early source of bacterial colonization for newborns. However, only a few small studies have investigated the composition of vaginal microbiota during labor. In this work, we analyzed vaginal swabs collected at 36 weeks gestation and at the onset of labor from 256 women participating in a randomized placebo-controlled study of probiotic supplementation for the prevention of atopic dermatitis in offspring. Although individuals' vaginal microbiota was stable over time, several bacterial families, which are characteristic of mixed community state type (CST) IV, were overrepresented in vaginal swabs sampled at labor. Alpha-diversity also tended to increase by between 36 weeks gestation and the onset of birth. In the majority of women, CST remained the same throughout the study. Among the women who switched their vaginal microbiota from one CST to another, approximately half shifted towards CST IV. Although CST IV is often associated with bacterial vaginosis, which in turn may lead to preterm birth, in our cohort this shift was not associated with self-reported vaginosis, preterm delivery or birthweight. Probiotic consumption did not alter vaginal microbiota.

Early gut mycobiota and mother-offspring transfer.

BACKGROUND: The fungi in the gastrointestinal tract, the gut mycobiota, are now recognised as a significant part of the gut microbiota, and they may be important to human health. In contrast to the adult gut mycobiota, the establishment of the early gut mycobiota has never been described, and there is little knowledge about the fungal transfer from mother to offspring. METHODS: In a prospective cohort, we followed 298 pairs of healthy mothers and offspring from 36 weeks of gestation until 2 years of age (1516 samples) and explored the gut mycobiota in maternal and offspring samples. Half of the pregnant mothers were randomised into drinking probiotic milk during and after pregnancy. The probiotic bacteria included Lactobacillus rhamnosus GG (LGG), Bifidobacterium animalis subsp. lactis Bb-12 and Lactobacillus acidophilus La-5. We quantified the fungal abundance of all the samples using qPCR of the fungal internal transcribed spacer (ITS)1 segment, and we sequenced the 18S rRNA gene ITS1 region of 90 high-quantity samples using the MiSeq platform (Illumina). RESULTS: The gut mycobiota was detected in most of the mothers and the majority of the offspring. The offspring showed increased odds of having detectable faecal fungal DNA if the mother had detectable fungal DNA as well (OR = 1.54, p = 0.04). The fungal alpha diversity in the offspring gut increased from its lowest at 10 days after birth, which was the earliest sampling point. The fungal diversity and fungal species showed a succession towards the maternal mycobiota as the child aged, with Debaryomyces hansenii being the most abundant species during breast-feeding and Saccharomyces cerevisiae as the most abundant after weaning. Probiotic consumption increased the gut mycobiota abundance in pregnant mothers (p = 0.01). CONCLUSION: This study provides the first insight into the early fungal establishment and the succession of fungal species in the gut mycobiota. The results support the idea that the fungal host phenotype is transferred from mother to offspring. TRIAL REGISTRATION: Clinicaltrials.gov NCT00159523.

Body fluid prediction from microbial patterns for forensic application.

The association of a DNA profile with a certain body fluid can be of essential importance in the evaluation of biological evidence. Several alternative methods for body fluid prediction have been proposed to improve the currently used presumptive tests. Most of them measure gene expression. Here we present a novel approach based on microbial taxonomic profiles obtained by standard 16S rRNA gene sequencing. We used saliva deposited on skin as a forensically relevant study model, but the same principle can be applied for predicting other bacteria rich body fluids. For classification we used standard pattern recognition based on principal component analysis in combination with linear discriminant analysis. A cross-validation of the experimental data shows that the new method is able to successfully classify samples from saliva deposited on skin and samples from pure skin in 94% of the cases. We found that there is a person-effect influencing the result, especially from skin, indicating that a reference sample of pure skin microbiota from the same person could improve accuracy. In addition the pattern recognition methods could be further optimized. Although there is room for improvement, this study shows the potential of microbial profiles as a new forensic tool for body fluid prediction.

Diet-Dependent Modular Dynamic Interactions of the Equine Cecal Microbiota.

Knowledge on dynamic interactions in microbiota is pivotal for understanding the role of bacteria in the gut. We herein present comprehensive dynamic models of the horse cecal microbiota, which include short-chained fatty acids, carbohydrate metabolic networks, and taxonomy. Dynamic models were derived from time-series data in a crossover experiment in which four cecum-cannulated horses were fed a starch-rich diet of hay supplemented with barley (starch intake 2 g kg-1 body weight per day) and a fiber-rich diet of only hay. Cecal contents were sampled via the cannula each h for 24 h for both diets. We observed marked differences in the microbial dynamic interaction patterns for Fibrobacter succinogenes, Lachnospiraceae, Streptococcus, Treponema, Anaerostipes, and Anaerovibrio between the two diet groups. Fluctuations and microbiota interactions were the most pronounced for the starch rich diet, with Streptococcus spp. and Anaerovibrio spp. showing the largest fluctuations. Shotgun metagenome sequencing revealed that diet differences may be explained by modular switches in metabolic cross-feeding between microbial consortia in which fermentation is linked to sugar alcohols and amino sugars for the starch-rich diet and monosaccharides for the fiber-rich diet. In conclusion, diet may not only affect the composition of the cecal microbiota, but also dynamic interactions and metabolic cross-feeding.