{
  "id": "metabolic-pathway",
  "title": "Metabolic Pathway",
  "lang": "en",
  "source_lang": "en",
  "summary": "A metabolic pathway is an organized sequence of biochemical reactions that transforms substances, transfers energy, and connects cellular functions within a wider metabolic network.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "Biochemical pathway"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/metabolic-pathway/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/metabolic-pathway/",
      "json": "https://aiwiki.page/zh/metabolic-pathway.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "A metabolic pathway is a connected sequence of chemical reactions through which a biological system converts starting substances into products. Most steps are accelerated by an enzyme, and the product of one step becomes a substrate for another. Pathways are functional subdivisions of metabolism, the complete set of chemical transformations that sustain living systems. They support energy conversion, the construction and breakdown of cellular components, nutrient assimilation, and the removal or reuse of metabolic products.\n\nA pathway is not usually an isolated assembly line. Its intermediates may enter several reactions, its activity depends on transport and regulation, and its boundaries often reflect how scientists organize a larger network. Understanding pathways therefore requires both molecular detail—what each enzyme does—and a systems perspective on the movement of matter, energy, and reducing equivalents. Pathway research connects fundamental biochemistry with physiology, evolution, medicine, agriculture, and industrial biotechnology."
    },
    {
      "heading": "Organization, boundaries, and pathway types",
      "text": "A pathway is commonly represented as a series of reactions, such as \\(A \\rightarrow B \\rightarrow C \\rightarrow D\\). The letters represent metabolites: substances produced, consumed, or transformed during metabolism. Each arrow represents a chemical conversion rather than necessarily a single enzyme molecule. Some enzymes catalyze several successive transformations, while some apparently simple conversions require a multienzyme complex. A complete description also identifies reaction partners, cofactors, cellular locations, and the conditions under which the reactions proceed. Omitting these details can make a pathway diagram convenient but chemically incomplete.\n\nPathways have several characteristic architectures. Linear pathways convert an initial substrate into a sequence of products; branched pathways distribute an intermediate among competing destinations. Cyclic pathways regenerate an acceptor molecule while processing incoming material, as occurs in the citric acid cycle. Spiral pathways repeat a similar sequence of transformations on a substrate whose size changes with each turn; fatty-acid degradation by beta-oxidation is a principal example. These categories describe reaction organization, not rigid biological compartments. A cycle may contain branches, and a nominally linear route may exchange intermediates extensively with other pathways.\n\nA second classification distinguishes catabolism, anabolism, and amphibolic metabolism. Catabolic pathways break down compounds and commonly conserve part of the released free energy in adenosine triphosphate (ATP) or reduced electron carriers. Anabolic pathways build compounds and generally require energy and reducing power. Amphibolic pathways serve both functions. The citric acid cycle oxidizes acetyl groups while supplying intermediates for the synthesis of amino acids, heme, and other compounds. Consequently, describing it solely as an energy-producing cycle overlooks its role in biosynthesis.\n\nPathway boundaries depend on the question being investigated. The conversion of glucose into lactate can be treated as one functional pathway, whereas another account separates glucose breakdown from the reaction that regenerates oxidized electron carrier. Similarly, a biosynthetic route may begin with an imported nutrient in one organism but with an internally synthesized precursor in another. A named pathway is therefore a useful biochemical unit, not necessarily an independently inherited or independently regulated module.\n\nA distinction must also be made between pathway presence and pathway operation. An organism may possess genes encoding all known reactions without carrying appreciable flux through them under a particular condition. Conversely, an apparently missing reaction may be supplied by an enzyme with an unrecognized secondary activity. Establishing a functioning pathway requires evidence that its component reactions connect in the relevant biological setting, rather than merely a list of plausible enzymes."
    },
    {
      "heading": "Energetics, redox balance, and chemical constraints",
      "text": "The direction and feasibility of a metabolic reaction depend on thermodynamics, not simply on the direction of an arrow in a textbook. The reaction Gibbs free-energy change is expressed as\n\n\\[\\Delta_r G=\\Delta_r G^{\\circ\\prime}+RT\\ln Q,\\]\n\nwhere \\(\\Delta_r G^{\\circ\\prime}\\) is a biochemical transformed standard reaction Gibbs energy under specified conditions, \\(R\\) is the gas constant, \\(T\\) is absolute temperature, and \\(Q\\) is the corresponding reaction quotient. The prime indicates a biochemical convention that accounts for specified conditions such as fixed pH. Cellular metabolite activities can differ greatly from standard conditions, so a positive standard value does not by itself establish that a reaction cannot proceed forward inside a cell.\n\nEnzymes lower the activation energy of reactions and increase the rates at which equilibrium can be approached; they do not change the reaction’s equilibrium position. A step near chemical equilibrium can reverse when substrate and product activities change. A strongly displaced step may function predominantly in one direction under physiological conditions, although it remains subject to thermodynamic law. Opposing pathways frequently use different enzymes to bypass such steps. Glucose synthesis, for example, is not simply glucose degradation run backward through every original enzyme.\n\nEnergetically unfavorable transformations can proceed when coupled to sufficiently favorable ones through a shared reaction mechanism. ATP-dependent phosphorylation is a common example, but ATP is not a universal source of usable energy independent of context. Its phosphorylation potential depends on ATP, ADP, phosphate, and relevant ionic conditions. Other coupling mechanisms include thioester chemistry and the movement of ions down an electrochemical gradient. In oxidative phosphorylation, electron-transfer reactions help establish a proton-motive force, and ATP synthase couples proton movement to ATP formation. Merely placing two reactions beside each other does not establish energetic coupling.\n\nMetabolic pathways must also balance oxidation and reduction. Oxidation–reduction reactions transfer electrons between substrates and carriers such as nicotinamide adenine dinucleotide (NAD) and its phosphorylated relative, NADP. NADH commonly supplies electrons to energy-conserving reactions, whereas NADPH commonly supports reductive biosynthesis and antioxidant systems. These roles are maintained through enzyme specificity and distinct intracellular redox states; they are not an absolute chemical division between two interchangeable molecules.\n\nEvery valid pathway must conserve atoms and electrical charge when all reactants and products are included. This requirement makes stoichiometry a powerful test of proposed mechanisms. A diagram that appears to create carbon, nitrogen, or reducing power without an input is incomplete or incorrect. Water, protons, inorganic phosphate, and electron carriers are sometimes omitted from simplified maps, but they become essential when calculating net reactions, energetic costs, and theoretical production yields."
    },
    {
      "heading": "Central pathways and the distribution of carbon",
      "text": "Central metabolism connects nutrient breakdown to energy conservation and the production of building blocks. In many organisms, glycolysis converts one six-carbon glucose molecule into two three-carbon pyruvate molecules through ten enzyme-catalyzed steps. In the conventional Embden–Meyerhof–Parnas pathway, two ATP molecules are invested and four are formed, giving a net yield of two ATP together with two NADH per glucose. This accounting describes conversion to pyruvate; the eventual products and total energetic yield depend on the reactions that follow.\n\nPyruvate occupies an important branch point. It can be converted into acetyl coenzyme A, carboxylated to oxaloacetate, or reduced to lactate, among other possibilities. Under appropriate conditions, acetyl coenzyme A enters the citric acid cycle. One conventional turn produces two carbon dioxide molecules, three NADH, one reduced flavin equivalent, and one GTP or ATP equivalent while regenerating oxaloacetate. These are net stoichiometric outputs; the two carbons released during a first turn do not simply correspond to the two carbons that have just entered as an acetyl group.\n\nThe pentose phosphate pathway diverts glucose 6-phosphate toward NADPH generation and sugar interconversion. Its oxidative reactions produce NADPH and a five-carbon sugar phosphate, while its nonoxidative reactions rearrange carbon skeletons between pentose phosphates and glycolytic intermediates. This flexibility permits different balances between reducing-power demand and the need for ribose phosphates used in nucleotide synthesis. Consequently, pathway activity cannot be inferred solely from how much glucose a cell consumes.\n\nBiosynthesis draws material away from central pathways. Citrate can support lipid synthesis, oxaloacetate can supply aspartate, and 2-oxoglutarate can supply glutamate. Withdrawal of cycle intermediates requires replenishment through anaplerotic reactions. Pyruvate carboxylase, for example, uses ATP and bicarbonate to convert pyruvate into oxaloacetate. Gluconeogenesis synthesizes glucose from suitable noncarbohydrate precursors through reactions shared with glycolysis and through distinct bypasses. In animals, ordinary even-chain fatty acids cannot provide net glucose through the conventional citric acid cycle because their acetyl units do not produce a net gain of cycle intermediates.\n\nOther major routes connect carbon and nitrogen metabolism. Amino-group transfer links amino-acid synthesis and degradation to carbon skeletons, while the urea cycle converts nitrogen into urea in mammals. In photosynthetic organisms, the Calvin cycle uses ATP and NADPH to assimilate inorganic carbon into organic compounds. These examples illustrate why no single pathway can represent “the metabolism” of all organisms: shared central reactions are embedded in different nutritional, physiological, and ecological systems."
    },
    {
      "heading": "Compartmentation, transport, and multiscale coordination",
      "text": "The location of a reaction is part of its biological identity. In eukaryotic cells, glycolysis occurs mainly in the cytosol, whereas much oxidative metabolism occurs in the mitochondrion. The mitochondrial inner membrane separates pools of metabolites and cofactors, sustains an electrochemical gradient, and contains transport proteins that determine access to matrix reactions. Two chemically identical metabolites in different compartments are therefore not automatically one freely mixing pool. A metabolic model that merges them can predict transformations that the intact cell cannot perform.\n\nTransport reactions connect compartments and may impose substantial energetic or kinetic constraints. Acetyl coenzyme A produced in mitochondria is not directly exported across the inner membrane to supply cytosolic fatty-acid synthesis; its carbon can instead be transferred through citrate. Likewise, cytosolic NADH does not simply cross the mitochondrial inner membrane. The malate–aspartate and glycerol phosphate shuttles transfer its reducing equivalents by coordinated reactions and transport processes. Their operation demonstrates that moving electrons between compartments is different from moving the original carrier molecule.\n\nPlant cells add further layers of organization. Chloroplasts contain photosynthetic carbon-assimilation reactions, while fatty-acid synthesis, nitrogen assimilation, and other processes have characteristic compartmental distributions. Photorespiration links chloroplasts, peroxisomes, and mitochondria through a sequence that recovers carbon from products of oxygenation by Rubisco. In many eukaryotes, peroxisomes also participate in fatty-acid oxidation, but their reaction organization and energy coupling differ from mitochondrial beta-oxidation. These differences matter when assigning enzymes or interpreting metabolite measurements.\n\nSpatial organization also occurs without membrane-bounded compartments. Enzymes can assemble into multienzyme complexes or associate with membranes and structural elements. Substrate channeling, in which an intermediate passes between catalytic sites without fully equilibrating with the surrounding solution, is firmly established in particular systems, including the internal ammonia transfer of tryptophan synthase. Broader claims that a collection of associated enzymes forms a channeling “metabolon” require direct evidence: physical proximity alone does not prove that intermediates bypass the bulk metabolite pool.\n\nCoordination extends beyond individual cells. In mammals, liver glucose production can supply tissues with different fuel requirements, and lactate released by one tissue can be consumed by another. In microbial communities, one organism’s fermentation products may become another’s substrates. Such exchanges create distributed metabolic functions without making the participants a single cell. The relevant pathway boundary may consequently lie at the level of an organ, a whole organism, or a community, depending on the process under study."
    },
    {
      "heading": "Regulation, flux control, and dynamic behavior",
      "text": "Pathway flux is the rate at which material passes through a reaction or sequence of reactions. It differs from metabolite concentration, which measures the size of a pool. A small pool can turn over rapidly, while a large pool can change slowly. At steady state, the rate of formation of an internal metabolite equals its rate of consumption, but both rates may be substantial. Steady state must therefore be distinguished from thermodynamic equilibrium, where there is no net driving force for the reaction concerned.\n\nRapid regulation often operates through enzyme activity. Substrates, products, and regulatory ligands can alter reaction rates, while allosteric regulation allows binding at one site to influence catalysis elsewhere in a protein. In mammalian glycolysis, phosphofructokinase-1 responds to indicators of energetic and metabolic state, including ATP, AMP, and fructose 2,6-bisphosphate. Feedback inhibition provides another common arrangement: an end product inhibits an earlier reaction, often near a branch’s entry point. This can limit unnecessary synthesis while preserving the availability of shared precursors.\n\nRegulation also occurs through reversible chemical modification and changes in enzyme abundance. Phosphorylation can alter enzyme activity over short timescales, whereas changes in gene expression adjust pathway capacity over longer periods. Hormonal signals coordinate these mechanisms across tissues. In mammalian liver, glucagon signaling favors glucose production and mobilization, while insulin signaling promotes storage and use of nutrients in appropriate physiological contexts. The response emerges from multiple targets, transport changes, and substrate supplies rather than a single molecular switch.\n\nThe idea that every pathway has one permanently “rate-limiting enzyme” is generally too simple. Metabolic control analysis, developed independently by Henrik Kacser and James Burns and by Reinhart Heinrich and Tom Rapoport in the 1970s, quantifies how changes in enzyme activities affect system behavior. A flux control coefficient can be written as\n\n\\[C_{E_i}^{J} = \\frac{\\partial \\ln J}{\\partial \\ln E_i},\\]\n\nwhere \\(J\\) is a specified steady-state flux and \\(E_i\\) is the activity of enzyme \\(i\\). Control can be distributed among several enzymes and transporters and can shift when conditions change. Under the assumptions of the classical framework, the relevant flux control coefficients sum to one.\n\nPathways can also display dynamic behavior that a static diagram cannot show. Delays, nonlinear enzyme responses, and feedback can produce oscillations, as observed in some glycolytic systems. Opposing pathways may operate simultaneously in a substrate cycle, consuming energy while increasing responsiveness or contributing to heat production. Such cycling is not necessarily an error, but its physiological role must be established in the particular system. Regulation is therefore concerned with timing and robustness as well as with maximizing chemical efficiency."
    },
    {
      "heading": "Evolution and metabolic diversity",
      "text": "Metabolic pathways evolve through changes in enzymes, regulation, transport, and cellular organization. Gene duplication can create an additional enzyme copy that later acquires altered substrate specificity or regulation. Enzyme promiscuity—the capacity to catalyze secondary reactions—provides another source of new transformations. A weak side activity can become biologically important if environmental conditions make its product useful and subsequent selection improves the reaction. Existing pathways can thereby acquire branches or new entry points without every component originating simultaneously.\n\nTwo influential ideas describe possible routes of pathway assembly. The retrograde hypothesis proposes that selection can extend a biosynthetic sequence backward from a required end product as environmental supplies diminish. Patchwork models emphasize recruitment of broadly active enzymes into new combinations, followed by specialization. These are explanatory frameworks rather than universal historical rules. Different pathways can have different origins, and present-day reaction order does not provide a direct record of the order in which their enzymes evolved.\n\nMetabolic capabilities also spread through horizontal gene transfer, especially among microorganisms. Acquisition of a gene cluster can supply several linked reactions, transport functions, and regulatory components. Conversely, gene loss can eliminate biosynthetic capacities when products are reliably supplied by hosts or neighboring organisms. Host-associated and symbiotic organisms often depend on such exchanges. A missing pathway is therefore not necessarily evidence of an unsuccessful organism; it may reflect a stable ecological dependence.\n\nThe diversity of energy and carbon metabolism is especially extensive in bacteria and archaea. Organisms can use different electron donors and acceptors, allowing growth under conditions where oxygen-dependent respiration is unavailable. Methanogenic archaea produce methane through specialized pathways, while nitrifying microorganisms oxidize reduced nitrogen compounds. Carbon assimilation likewise has several solutions. In addition to the Calvin cycle, organisms use routes such as the reverse citric acid cycle and the reductive acetyl-coenzyme A pathway. These systems differ in energetic requirements, enzyme sensitivity, and ecological distribution.\n\nConservation and diversity coexist. Glycolytic reactions are widespread, but alternative routes such as the Entner–Doudoroff pathway distribute glucose carbon and reducing equivalents differently. Some enzymes with similar functions have distinct evolutionary origins, whereas related enzymes may catalyze different reactions. Pathway reconstruction therefore requires more than identifying a familiar gene name. Comparative analysis combines sequence relationships, experimentally established activities, gene neighborhoods, physiology, and environmental context to distinguish inherited similarity from independent biochemical solutions."
    },
    {
      "heading": "Discovery and experimental reconstruction",
      "text": "The pathway concept developed through experiments showing that biological transformations could be resolved into reproducible chemical steps. In 1897, Eduard Buchner demonstrated fermentation in cell-free yeast extracts, establishing that intact living cells were not required for the chemical conversion itself. This helped make enzyme-catalyzed reactions experimentally accessible. Work by Gustav Embden, Otto Meyerhof, Jakub Parnas, and others subsequently clarified glycolytic intermediates and reaction sequences. Hans Krebs and Kurt Henseleit described the urea cycle in 1932, and Krebs and William Johnson proposed the citric acid cycle in 1937.\n\nEarly reconstruction relied on isolating intermediates, measuring reaction products, and observing the effects of inhibitors or missing nutrients. George Beadle and Edward Tatum’s 1941 studies of Neurospora crassa linked genetic changes to specific biochemical requirements, helping establish that genes govern identifiable steps in metabolic processes. A mutant that grows only when supplied with a particular compound can reveal a biosynthetic defect. However, nutrient rescue alone does not specify an enzyme mechanism, because supplementation can bypass several possible failures.\n\nIsotope tracing made it possible to follow atoms rather than merely compare initial and final compounds. The carbon-14 experiments of Melvin Calvin, Andrew Benson, James Bassham, and colleagues helped establish the sequence of photosynthetic carbon assimilation. Contemporary studies commonly use stable isotopes, especially carbon-13 and nitrogen-15. Labeling glucose at selected carbon positions, then measuring label distributions in downstream compounds, can distinguish competing routes that produce the same final metabolite. Interpretation requires accounting for reversible reactions, mixing of pools, and the duration of labeling.\n\nModern metabolomics measures many small molecules in parallel, frequently using chromatography coupled to mass spectrometry, or nuclear magnetic resonance. Rapid sampling and quenching are important because some metabolite pools change within seconds. Measurements also face chemical challenges: different compounds have different detection efficiencies, some metabolites are unstable, and compounds with the same mass may require additional separation or structural evidence. A detected signal is not automatically a securely identified metabolite.\n\nStrong pathway reconstruction combines complementary evidence. Purified-enzyme experiments establish catalytic capability; genetic perturbations test biological dependence; isotope tracing establishes atom flow; and cellular measurements show whether substrates, products, and cofactors occur in compatible locations and quantities. Each method has limitations. An enzyme may catalyze a reaction in vitro at concentrations never encountered in vivo, while deletion of a gene may activate compensatory routes. The most persuasive assignment explains chemistry, genetics, localization, and flux within one consistent account."
    },
    {
      "heading": "Network models and computational analysis",
      "text": "A metabolic network can be represented mathematically by a stoichiometric matrix, usually denoted \\(S\\). Each column corresponds to a reaction and each row to a metabolite, with negative coefficients for consumption and positive coefficients for production. If \\(\\mathbf{x}\\) is the vector of metabolite concentrations and \\(\\mathbf{v}\\) the vector of reaction rates, then an idealized fixed-volume system satisfies\n\n\\[\\frac{d\\mathbf{x}}{dt}=S\\mathbf{v}.\\]\n\nAdditional terms are needed when dilution, volume changes, or exchanges are not already represented. This formulation makes shared intermediates explicit and connects individual pathways into a common accounting system.\n\nAt metabolic steady state, internal balances satisfy \\(S\\mathbf{v}=0\\). Flux balance analysis combines this relation with bounds on uptake, secretion, and reaction direction, then searches for flux distributions that optimize a specified objective. A common microbial objective is production of a biomass mixture representing cellular material. The method can analyze large networks without specifying every enzyme’s kinetic parameters. Its predictions depend on the reconstruction, constraints, and objective, however, and an optimal mathematical solution is not automatically a unique or experimentally realized physiological state.\n\nGenome-scale reconstructions connect reactions to genes and proteins. Gene–protein–reaction rules distinguish enzyme complexes, which require several gene products, from alternative enzymes that can perform the same reaction. Reconstructions must also assign compartments, exchange reactions, transport processes, and biomass requirements. Gaps may be filled computationally by adding reactions that restore a required capability, but such additions remain hypotheses until supported. Incorrect reaction reversibility or an unbalanced transport step can create artificial cycles that appear to generate ATP without a legitimate energy source.\n\nOther modeling approaches address questions that steady-state stoichiometry alone cannot resolve. Kinetic models describe rates as functions of metabolite concentrations and enzyme properties, enabling analysis of transients and regulation. Carbon-13 metabolic flux analysis estimates fluxes by fitting isotope-labeling patterns to atom-resolved reaction models. Elementary flux modes identify minimal steady-state sets of reactions under specified definitions, while thermodynamic constraints can exclude flux directions incompatible with feasible chemical potentials. These methods differ in data requirements and in the kinds of uncertainty they expose.\n\nPathway databases and computational maps are valuable organizing tools, but they combine different kinds of evidence. An experimentally characterized reaction in one species may be inferred in another through sequence similarity. A reference map may also combine reactions that never coexist in a single organism or compartment. Careful use therefore distinguishes a general biochemical possibility from a species-specific reconstruction and from a measured condition-specific flux state. Model disagreement can be informative when it identifies a missing transporter, an incorrect enzyme assignment, or an unrecognized metabolic capability."
    },
    {
      "heading": "Disease mechanisms and metabolic engineering",
      "text": "A pathway defect can have several consequences at once: accumulation of upstream substrates, depletion of downstream products, diversion into alternative reactions, and changes in energy or redox balance. The effect depends on tissue distribution, residual enzyme activity, nutrient supply, and compensatory routes. In phenylketonuria, impaired phenylalanine hydroxylation disrupts conversion of phenylalanine to tyrosine and can cause phenylalanine accumulation. Defects in urea-cycle enzymes can impair nitrogen disposal and produce hyperammonemia. These examples show why an enzyme’s local reaction must be interpreted within the larger physiological network.\n\nAcquired disease can also alter pathway use without removing a single reaction. Many cancers exhibit increased glucose uptake and lactate production despite the availability of oxygen, a pattern associated with the Warburg effect. This does not imply that all cancer cells lack mitochondrial respiration; metabolic behavior varies with tumor type, genotype, oxygen supply, and interactions with surrounding cells. Proliferating cells also need carbon and nitrogen for nucleotides, proteins, and membranes, making biosynthetic routing as important as ATP production. Measurements of metabolite abundance alone cannot establish the full direction or magnitude of these flux changes.\n\nPathways provide targets for medicines because particular reactions may be essential or selectively vulnerable. Statins inhibit HMG-CoA reductase in cholesterol biosynthesis, while methotrexate interferes with folate-dependent metabolism through inhibition of dihydrofolate reductase. Antimicrobial selectivity can arise when pathogens synthesize a required compound that their hosts obtain differently, or when their enzymes have distinct molecular properties. The relationship between target inhibition and clinical effect nevertheless depends on exposure, tissue distribution, compensation, and other biological factors; a plausible pathway target is not itself evidence of therapeutic efficacy.\n\nIn metabolic engineering, pathways are deliberately modified to increase production of fuels, chemicals, pharmaceuticals, or food ingredients. Strategies include increasing precursor supply, redirecting competing branches, balancing cofactors, and improving product export. Engineered yeast production of artemisinic acid, a precursor used in semisynthetic artemisinin manufacture, illustrates the combination of native isoprenoid metabolism with introduced plant-derived reactions. Installing the desired enzymes was only part of the task: sufficient carbon flow and compatible expression of the pathway also had to be achieved.\n\nEngineering is constrained by the same chemistry as natural metabolism. A route with a favorable theoretical carbon yield may demand excessive ATP, produce a toxic intermediate, or require more NADPH than the host can supply. Increasing one enzyme’s abundance can worsen an imbalance rather than increase product formation. Industrial development therefore links pathway design to fermentation conditions, transport, genetic stability, and downstream recovery. At community scale, engineered consortia can divide transformations among organisms, but their performance additionally depends on metabolite exchange and the stability of the participating populations."
    }
  ],
  "links": [
    "chemical-reaction",
    "enzyme",
    "metabolism",
    "biochemistry",
    "citric-acid-cycle",
    "beta-oxidation",
    "adenosine-triphosphate",
    "amino-acid",
    "glucose",
    "thermodynamics",
    "gibbs-free-energy",
    "activation-energy",
    "chemical-equilibrium",
    "electrochemical-gradient",
    "oxidative-phosphorylation",
    "atp-synthase",
    "redox-reaction",
    "stoichiometry",
    "glycolysis",
    "pentose-phosphate-pathway",
    "nucleotide",
    "gluconeogenesis",
    "urea-cycle",
    "calvin-cycle",
    "eukaryote",
    "mitochondrion",
    "chloroplast",
    "allostery",
    "feedback",
    "gene-expression",
    "horizontal-gene-transfer",
    "bacteria",
    "archaea",
    "methane",
    "isotope",
    "metabolomics",
    "chromatography",
    "mass-spectrometry",
    "nuclear-magnetic-resonance",
    "matrix"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}