{
  "id": "microbiome",
  "title": "Microbiome",
  "lang": "en",
  "source_lang": "en",
  "summary": "A microbiome is a microbial community, its genetic material, and its ecological setting, studied for its roles in organisms, environments, and biogeochemical processes.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "Microbial biome"
  ],
  "generated_at": "2026-11-02",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/microbiome/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/microbiome/",
      "json": "https://aiwiki.page/zh/microbiome.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A microbiome is the assemblage of microorganisms associated with a particular environment, considered together with their genetic material, activities, and ecological setting. Microbiomes occur in soils, oceans, sediments, buildings, and the bodies of plants and animals. Depending on the definition adopted, the term may emphasize either the collective genomes of a microbial community or the broader ecological system comprising its organisms, molecules, and local environmental conditions. The related term microbiota refers more specifically to the organisms themselves.\n\nMicrobiomes influence processes ranging from digestion and immune development to soil fertility and the global carbon cycle. Their members transform substances, exchange metabolites, compete for resources, and alter the conditions experienced by other organisms. Advances in DNA sequencing have made it possible to investigate communities that cannot be adequately characterized by conventional cultivation alone. Microbiome research consequently links microbiology, ecology, medicine, agriculture, and Earth-system science, while requiring careful distinctions between community composition, biological activity, and demonstrated causal effects."
    },
    {
      "heading": "Definitions and development of the field",
      "text": "The meaning of microbiome has changed with the development of microbial ecology. In a frequently cited 1988 formulation, the microbial ecologists John Whipps, Karen Lewis, and Richard Cook used the term for a characteristic microbial community occupying a reasonably well-defined habitat with distinctive physicochemical properties. Later usage often emphasized the collective genetic information of associated microbes. Joshua Lederberg helped popularize the concept in discussions of human-associated microbial communities around the beginning of the twenty-first century, but the term predates those discussions. Contemporary definitions commonly distinguish the microbial community from its “theatre of activity”: structural components, metabolites, genetic material, and environmental conditions that participate in its functioning.\n\nThese distinctions matter in practice. A list of bacterial species from a stool sample describes only part of an intestinal microbiota. The community’s combined genomes constitute its metagenome, whereas its expressed genes, proteins, metabolites, and interactions reveal different aspects of its activity. A sample can contain DNA from dead cells or material transported from elsewhere, so the detection of a sequence does not necessarily establish that a living organism is resident or active. Likewise, a microbiome is not automatically a sharply bounded natural unit. Its boundaries depend partly on the question being investigated: a leaf surface, an intestinal segment, a soil aggregate, or an entire drainage system can each be a relevant scale.\n\nThe intellectual foundations extend well before modern sequencing. Seventeenth-century microscopy revealed microorganisms in dental material and water. During the nineteenth century, Louis Pasteur and other investigators established microbial roles in fermentation and disease. Sergei Winogradsky and Martinus Beijerinck subsequently helped establish environmental microbiology through studies of microbial transformations and enrichment cultures. Their work demonstrated that microorganisms could be understood through their chemical activities and ecological requirements, rather than solely as isolated pathogens. Research on rumen fermentation, soil nutrient transformations, and aquatic microbial food webs likewise anticipated many present-day microbiome questions.\n\nA major transition followed the use of ribosomal RNA sequences to infer evolutionary relationships. Carl Woese and George Fox’s 1977 work established the distinctiveness of the lineage now called archaea, illustrating how molecular evidence could overturn classifications based primarily on appearance and physiology. Environmental sequencing later revealed extensive diversity missing from laboratory collections. The Human Microbiome Project, launched in 2007, and large environmental initiatives expanded standardized community surveys. The field has since moved beyond inventories toward strain-resolved analysis, cultivation of previously inaccessible organisms, and experiments designed to connect particular microbial activities with host or environmental outcomes."
    },
    {
      "heading": "Community membership and spatial organization",
      "text": "Microbiomes can contain bacteria, archaea, microscopic eukaryotes, and viruses. Eukaryotic members include fungi, protists, and, depending on the study’s scope, microscopic animals. Viral communities include viruses of animals, plants, fungi, and microbial hosts; bacteriophages infect bacteria and can influence bacterial abundance, evolution, and gene exchange. Some definitions restrict microbiota to cellular organisms while including viruses within the broader microbiome. Explicitly stating the scope of a study is therefore more informative than assuming that every microbiome dataset covers the same biological groups.\n\nMembership can be described at several taxonomic resolutions. A community might appear stable at the level of genus while undergoing substantial replacement of strains. Strains assigned to the same species can differ in nutrient use, host interaction, toxin production, or resistance to antimicrobial compounds. Conversely, organisms that are only distantly related may perform similar biochemical tasks. This partial functional redundancy helps explain why two communities can differ taxonomically yet carry out comparable processes. Redundancy is rarely complete, however: organisms with similar metabolic capacities may differ in growth rate, environmental tolerance, or the conditions under which they express those capacities.\n\nSpatial structure is often as important as membership. In a biofilm, cells attach to a surface or one another and inhabit a matrix containing polysaccharides, proteins, and extracellular DNA. Consumption of nutrients and restricted diffusion establish gradients across distances of micrometres to millimetres. An oxygenated surface may lie close to an anoxic interior, allowing different forms of metabolism to coexist within a seemingly uniform sample. Comparable spatial partitioning occurs within soil aggregates, marine particles, and intestinal mucus. Sampling an entire surface or homogenizing material can conceal these localized arrangements.\n\nCommunity interactions include competition, predation, parasitism, and exchanges that benefit one or more participants. In cross-feeding, one organism consumes a product released by another. In syntrophy, a metabolic transformation may become energetically feasible because a partner continually removes a reaction product. Hydrogen-consuming methanogens, for example, can enable fermentative partners to carry out reactions that would otherwise yield too little usable energy. Other organisms inhibit neighbours with acids, antimicrobial compounds, or contact-dependent mechanisms.\n\nThe balance of these interactions is conditional rather than fixed. A molecule that benefits a neighbour under one nutrient regime may become irrelevant under another, and an organism that is harmless in one location can cause disease in another. Microbiome membership therefore does not divide neatly into universally beneficial and universally harmful species. Biological effects depend on strain properties, abundance, activity, location, and the state of the surrounding community or host."
    },
    {
      "heading": "Assembly, succession, and stability",
      "text": "Microbiome assembly reflects the interaction of dispersal, environmental selection, biological interactions, and chance. Organisms must first reach a habitat, whether through water, air, food, physical contact, or transfer between generations. Establishment then depends on local conditions such as temperature, pH, salinity, oxygen availability, and accessible nutrients. A microorganism that repeatedly enters a habitat may remain only a transient visitor if it cannot reproduce there. Conversely, a successfully established population can persist even when its abundance is too low for routine detection.\n\nThe order of arrival can influence later community development through priority effects. Early colonists may consume limiting resources, occupy attachment sites, or change local chemistry. They can thereby exclude later arrivals or prepare conditions that enable them to establish. For example, organisms consuming oxygen can create conditions suitable for anaerobic populations. Such processes contribute to ecological succession on newly exposed surfaces, in developing animal guts, and during the decomposition of plant material. Assembly is not entirely deterministic: random losses and gains matter especially when populations are small or dispersal events are infrequent.\n\nMicrobiome diversity has several distinct meanings. Alpha diversity describes diversity within a sample, whereas beta diversity describes differences among samples. Species richness counts detected taxa; other measures also account for their relative abundances. The Shannon index, for example, is\n\n\\[H=-\\sum_{i=1}^{S}p_i\\ln p_i,\\]\n\nwhere \\(S\\) is the number of taxa and \\(p_i\\) is the proportion assigned to taxon \\(i\\). The index depends on sampling and analytical choices, and it does not directly measure ecosystem benefit. A diverse soil community and a healthy vaginal community can have very different abundance patterns. No single diversity value provides a universal assessment of microbiome quality.\n\nStability also has multiple dimensions. Resistance describes how little a community changes during a disturbance, while resilience concerns its recovery afterward. Composition and function need not recover together. A community can regain a process such as organic-matter decomposition without returning to its original species composition, or recover familiar taxa while retaining altered activity. Disturbances include drought, flooding, dietary changes, infection, and exposure to an antibiotic. Their effects depend on intensity, duration, and previous community history.\n\nEvolution occurs alongside ecological change. Mutations, natural selection, and genetic drift alter populations, while horizontal gene transfer can move traits between lineages. Plasmids and other mobile genetic elements can distribute antimicrobial resistance or metabolic capabilities without the wholesale replacement of species. As a result, a taxonomically stable microbiome may still change genetically and functionally over time."
    },
    {
      "heading": "Sampling, sequencing, and functional measurement",
      "text": "A microbiome study begins with decisions about what, where, and when to sample. Stool, saliva, tissue biopsies, seawater filters, and soil cores capture different fractions of microbial habitats. Stool is convenient for studying the intestinal community but does not fully represent the small intestine or organisms associated with the intestinal wall. In environmental studies, sample depth, season, particle size, and storage can strongly affect results. Freezing, preservatives, oxygen exposure, and delays before processing can alter measurable community properties, making collection protocols part of the scientific measurement rather than merely logistical details.\n\nMarker-gene sequencing identifies organisms using selected genetic regions. Bacterial and archaeal surveys commonly target the 16S ribosomal RNA gene; fungal surveys often target the internal transcribed spacer region. These methods usually involve polymerase chain reaction amplification and provide an efficient account of community composition. They nevertheless introduce biases through primer choice, amplification efficiency, and variation in marker-gene copy number. They also offer limited resolution for many closely related strains and cannot by themselves establish the full metabolic capabilities of detected organisms.\n\nShotgun metagenomics sequences DNA across the sampled material rather than a single marker. Reads can be compared with reference databases or assembled into longer sequences, from which metagenome-assembled genomes may be reconstructed. This approach can detect functional genes, distinguish some strains, and recover previously uncharacterized organisms. Its performance depends on sequencing depth, community complexity, and reference coverage. Assembly may combine or separate closely related populations imperfectly, and a reconstructed genome is not necessarily equivalent to a cultured, experimentally characterized organism.\n\nFunctional measurements add complementary evidence. Metatranscriptomics measures community RNA and can indicate which genes are being expressed. Metaproteomics measures community proteins, while metabolomics examines small molecules. Mass spectrometry, microscopy, flow cytometry, and stable-isotope tracing can connect chemical transformations with particular populations. Cultivation remains essential for testing physiological properties and interactions. Anaerobic culture systems, high-throughput cultivation, and defined experimental communities complement sequencing rather than being superseded by it.\n\nInterpretation must account for the difference between relative and absolute abundance. If sequencing reports \\(p_i=n_i/\\sum_j n_j\\), a taxon’s proportion can increase even when its absolute abundance \\(n_i\\) remains unchanged, simply because other populations decline. Cell counts, quantitative PCR, or internal standards can help resolve this ambiguity. Contamination is especially consequential in low-biomass samples, where DNA from reagents or handling can rival the biological signal. Negative controls, mock communities, batch-aware analysis, and independent replication are consequently central safeguards. Detection of DNA alone does not establish viability, local growth, or biological activity."
    },
    {
      "heading": "Human body habitats and development across life",
      "text": "The human microbiome consists of distinct communities associated with different body sites rather than one uniform assemblage. The large intestine contains dense populations adapted to low-oxygen conditions and complex nutrient supplies. The mouth contains multiple habitats, including teeth, tongue, saliva, and gingival crevices. Skin communities vary with moisture, temperature, exposure, and secretions: oily, moist, and dry sites support different organisms. The vagina and respiratory tract have their own ecological conditions. Microbial composition therefore varies substantially even between adjacent anatomical locations.\n\nIn the colon, many abundant organisms belong to bacterial groups traditionally called Firmicutes and Bacteroidetes, now often designated Bacillota and Bacteroidota in revised classifications. These broad categories encompass organisms with markedly different functions and are insufficient by themselves to characterize health. Archaea such as Methanobrevibacter smithii occur in some individuals and consume hydrogen during methane production. Fungi and viruses form additional, less completely characterized components. The oral cavity illustrates the importance of spatial organization: tooth surfaces provide persistent attachment sites, whereas many other body surfaces continually shed cells.\n\nExtensive colonization develops during and after birth. Delivery circumstances, maternal contact, feeding, household exposures, and antibiotic use influence early community development. Breast milk supplies both microorganisms and human milk oligosaccharides, which particular infant-associated bacteria can use. Some bifidobacteria possess specialized systems for metabolizing these compounds. Claims of a stable resident placental microbiome in healthy pregnancy have not been consistently supported by contamination-controlled studies; microbial detection in such low-biomass material requires particular caution.\n\nDuring childhood, the introduction of solid foods, changing physiology, and repeated environmental exposure reshape microbial communities. Adult communities can retain individualized features over time while still responding to changes in diet, medication, infection, or living conditions. Microbiomes also change in older populations, but chronological age is not an isolated explanation: diet, residence, illness, and medication use frequently vary alongside it. Longitudinal sampling is needed to separate persistent individual characteristics from short-term fluctuations.\n\nHuman microbiome surveys differ in geographic coverage, and early sequencing studies disproportionately represented industrialized populations. Studies across diverse settings have documented differences associated with subsistence, diet, sanitation, and other environmental conditions. These findings do not establish a single ancestral or universally optimal human microbiome. Host genetics contributes to the abundance of some organisms, but shared environment and exposure also have substantial effects. Family resemblance can therefore arise through transmission and common living conditions as well as inherited host traits."
    },
    {
      "heading": "Metabolic and immune relationships with hosts",
      "text": "Host-associated microorganisms expand the range of chemical transformations occurring within an animal. In the human colon, microbial enzymes break down components of dietary fibre and other carbohydrates that host enzymes do not fully digest. Fermentation produces short-chain fatty acids, including acetate, propionate, and butyrate. These compounds can be absorbed and used by the host; butyrate is an important energy substrate for colonocytes. Their effects also involve receptors, gene regulation, and local chemical conditions, so their biological significance cannot be inferred solely from the abundance of one producing species.\n\nMicrobial transformations extend beyond dietary carbohydrate. Intestinal organisms modify bile acids, metabolize amino acids, and produce compounds derived from dietary and host molecules. They can synthesize certain vitamins, although synthesis does not automatically imply that a product is available to the host in a nutritionally meaningful quantity. Microbes can also alter medicines. Some intestinal bacteria convert digoxin to a less active form, with the effect depending on strain-level genetic capabilities and environmental conditions. Such examples demonstrate why taxonomic surveys alone do not fully predict drug metabolism.\n\nThe microbiome interacts extensively with the immune system. Epithelial barriers, mucus, antimicrobial molecules, and secreted antibodies help control microbial access to host tissues. Signals detected by innate immunity influence immune development and activity. Components of adaptive immunity, including immunoglobulin A and regulatory T cells, contribute to regulated interactions with commensal organisms. Experimental work in germ-free animals has shown that microbial exposure affects intestinal anatomy and immune characteristics, although these models differ substantially from conventionally colonized humans.\n\nEstablished communities can provide colonization resistance: protection against invading organisms through resource competition, production of inhibitory substances, and modification of the habitat. Bile-acid transformations offer one mechanistic example relevant to Clostridioides difficile. Certain microbial products inhibit stages of its growth, while antibiotic disruption can remove competitors and alter the chemical environment in ways that permit expansion. Colonization resistance is a community property rather than an invariant attribute of any single species, and pathogens may exploit nutrients or conditions created during inflammation.\n\nHost–microbe relationships also involve costs and conflicts. An organism that consumes intestinal mucus may participate in normal nutrient cycling but contribute to barrier erosion under particular conditions. A microbial product can have different effects depending on its concentration and the tissue exposed to it. The holobiont concept describes a host together with its associated microorganisms, but does not establish that all partners share interests or evolve as a single integrated unit. Transmission routes, host control, and conflicts among microbial populations influence how closely their evolutionary trajectories align."
    },
    {
      "heading": "Disease associations and causal evidence",
      "text": "Microbiome differences have been reported in many conditions, including inflammatory bowel disease, metabolic disorders, colorectal cancer, and infectious diseases. The evidential meaning of these associations varies. A disease can change diet, intestinal transit, immune activity, oxygen availability, and medication exposure, each of which can reshape microbial communities. Consequently, an organism associated with disease may contribute to pathology, benefit from the altered habitat, or reflect another causal factor. Statistical separation between patient and comparison groups does not, by itself, establish a disease mechanism.\n\nThe term dysbiosis is widely used for a disturbed microbial community, but it lacks a single operational definition. In different studies it may mean loss of particular organisms, expansion of opportunistic populations, altered diversity, or disrupted metabolic activity. Without a specified measurement and biological interpretation, the term can merely restate that a community differs between groups. Similarly, broad ratios between major bacterial groups are not reliable universal indicators of health. Diagnostic interpretation requires evidence appropriate to the intended population and clinical setting.\n\nStronger causal evidence comes from complementary approaches. Longitudinal studies can establish whether a change precedes an outcome. Controlled perturbations test whether altering a microbial exposure changes a phenotype. Germ-free animals and animals colonized with defined communities allow experiments that are difficult or impossible in humans, while microbial mutants can identify particular genes or metabolites involved. Nevertheless, transferring a human community into a mouse also transfers it into a different diet, anatomy, immune system, and ecological setting. A phenotype reproduced in that model supports a mechanism under the experimental conditions, not automatic equivalence to human disease.\n\nRecurrent C. difficile infection provides an important example of clinically demonstrated microbiome-directed treatment. Disruption following antibiotic exposure can impair resistance to this pathogen. Trials of fecal microbiota transplantation have shown that introducing screened donor-derived microbial material can prevent recurrence in selected patients. In the United States, the Food and Drug Administration approved the rectally administered product Rebyota in 2022 and the oral product Vowst in 2023 for prevention of recurrent infection in adults following antibacterial treatment for recurrent C. difficile infection. These indications are specific and do not establish efficacy for unrelated conditions.\n\nMicrobial treatments also present safety challenges because biological material may transmit pathogens or antimicrobial-resistance determinants. Donor screening, manufacturing controls, and surveillance are therefore integral to clinical use. For many other proposed microbiome interventions, evidence remains dependent on the particular product, strain, population, and outcome studied. Claims about effects on mood, body weight, or general wellness cannot be inferred simply from a plausible pathway or an association. Randomized controlled trials, mechanistic experiments, and reproducible measurements address different parts of the causal question and are most informative when their findings converge."
    },
    {
      "heading": "Environmental microbiomes and planetary processes",
      "text": "Environmental microbiomes mediate transformations essential to ecosystems. Soil communities decompose organic matter, release nutrients, interact with roots, and contribute to the formation and breakdown of organic compounds associated with mineral surfaces. These processes influence soil structure and carbon storage. Microbial activity does not simply accelerate carbon loss: microbial residues can also become stabilized in soil. The net outcome depends on substrate chemistry, moisture, temperature, mineral interactions, and access to nutrients, making the relationship between microbial diversity and carbon persistence context-dependent.\n\nMicroorganisms carry out major steps of the nitrogen cycle. Nitrogen fixation converts atmospheric nitrogen into biologically accessible forms, while nitrification oxidizes reduced nitrogen compounds and denitrification returns oxidized nitrogen to gaseous forms. Anaerobic ammonium oxidation provides another route to nitrogen gas. These transformations may occur in neighbouring microenvironments within the same soil or sediment. Some pathways produce nitrous oxide, a greenhouse gas, whereas others consume it. Understanding the distribution and expression of relevant genes helps explain nitrogen retention and loss, but gene detection alone cannot specify process rates.\n\nMarine microbiomes connect primary production with the recycling and export of organic matter. Cyanobacteria, including Prochlorococcus and Synechococcus, contribute substantially to photosynthesis in many ocean regions. Heterotrophic microorganisms consume dissolved and particulate organic matter, returning nutrients to circulation. Viruses lyse host cells and redirect material into dissolved pools, while microbes associated with sinking particles can influence how much carbon reaches deeper water. In oxygen-depleted waters and sediments, communities use alternative electron acceptors and carry out transformations unavailable to organisms restricted to aerobic respiration.\n\nMicrobial activity also affects methane. Methanogenic archaea generate methane in anoxic environments such as wetlands, sediments, and animal digestive systems. Other microorganisms oxidize methane before it reaches the atmosphere, using oxygen or, in some settings, alternative electron acceptors through microbial partnerships. The balance between methane production and consumption is sensitive to water level, temperature, and substrate availability. In thawing permafrost, newly available organic matter and changing hydrology alter microbial pathways, contributing to feedbacks between ecosystems and climate change.\n\nPlant-associated microbiomes occupy roots, leaves, internal tissues, and surrounding soil. The rhizosphere, the soil region influenced by roots, receives sugars, organic acids, and other root-derived compounds that select for particular populations. Some microbes improve nutrient acquisition or suppress plant pathogens, while others cause disease. Reef-building corals likewise associate with bacteria, archaea, viruses, and photosynthetic microbial eukaryotes, but these partners have distinct functions and should not be treated as interchangeable. Across environmental systems, microbial activity connects local interactions at microscopic scales with measurable changes in nutrient availability, productivity, and atmospheric chemistry."
    },
    {
      "heading": "Biotechnology, intervention, and research governance",
      "text": "Microbial communities have long been used in food production, waste treatment, and agriculture, even before the term microbiome became widespread. Cheese, sourdough, and other fermented foods depend on communities whose members alter acidity, texture, aroma, and resistance to spoilage. Industrial control requires attention not only to the presence of desired organisms but also to their succession and interactions. Two fermentations with similar ingredient lists can diverge because of differences in starter populations, temperature, salt concentration, oxygen exposure, or contamination.\n\nWastewater treatment illustrates community engineering at large scale. Activated-sludge systems retain diverse microbial populations that transform organic pollutants and nutrients. Operating conditions select for functions such as nitrification, denitrification, and biological phosphorus removal. Anaerobic digesters use interacting communities to convert organic material into biogas, with fermentative organisms, syntrophic partners, and methanogenic archaea carrying out successive steps. Failure can arise when one part of the metabolic chain outpaces another, causing inhibitory compounds to accumulate. Community monitoring therefore complements measurements of chemical performance.\n\nAgricultural applications include microbial inoculants, disease-suppressive communities, and strategies intended to alter root-associated populations. Their performance in controlled experiments does not necessarily predict field performance. Introduced organisms must survive storage, reach the target habitat, compete with resident populations, and remain active under variable soil and climatic conditions. bioremediation similarly depends on matching microbial capabilities with environmental constraints. A community may contain genes for pollutant degradation yet fail to remove the compound if oxygen, nutrients, accessibility, or suitable partners are limiting.\n\nDesigned communities, sometimes called synthetic microbial consortia, provide a way to test and exploit defined interactions. Researchers can assemble strains with complementary metabolic functions and track how community performance changes when a member or gene is removed. Genome editing and controlled bioreactors extend these possibilities, but engineering an open ecosystem is more difficult than controlling a closed vessel. Dispersal, competition, evolutionary change, and transfer of genetic material can alter the behaviour of introduced populations. Persistence and ecological effects therefore require separate evaluation from the intended biochemical function.\n\nResearch governance includes reproducibility, equitable sampling, and responsible handling of human-associated data. Sequencing can capture host genetic material alongside microbial DNA, creating data privacy concerns. Comparisons across studies also depend on transparent collection protocols, reference databases, analytical software, and reporting standards. Commercial and clinical tests require validation for their specific purposes; a technically accurate microbial profile is not automatically a clinically meaningful diagnosis. In environmental and human research alike, a central practical challenge is to connect a reproducible measurement with a defined mechanism and an outcome that can be independently verified."
    }
  ],
  "links": [
    "microorganism",
    "carbon-cycle",
    "dna-sequencing",
    "microbiology",
    "genome",
    "louis-pasteur",
    "ribosomal-rna",
    "archaea",
    "bacteria",
    "eukaryote",
    "virus",
    "bacteriophage",
    "species",
    "biofilm",
    "diffusion",
    "metabolism",
    "temperature",
    "ph",
    "antibiotic",
    "mutation",
    "natural-selection",
    "genetic-drift",
    "horizontal-gene-transfer",
    "plasmid",
    "polymerase-chain-reaction",
    "metagenomics",
    "proteomics",
    "metabolomics",
    "mass-spectrometry",
    "flow-cytometry",
    "isotope",
    "enzyme",
    "immune-system",
    "innate-immunity",
    "adaptive-immunity",
    "regulatory-t-cell",
    "inflammation",
    "cancer",
    "randomized-controlled-trial",
    "ecosystem",
    "nitrogen-fixation",
    "greenhouse-gas",
    "cyanobacteria",
    "methane",
    "permafrost",
    "climate-change",
    "genome-editing",
    "data-privacy"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}