{
  "id": "homology-biology",
  "title": "Homology (biology)",
  "lang": "en",
  "source_lang": "en",
  "summary": "Homology is the correspondence between biological features inherited from a common ancestor, underlying comparisons of anatomy, development, genes, and genomes.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "Biological homology",
    "evolutionary homology"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/homology-biology/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/homology-biology/",
      "json": "https://aiwiki.page/zh/homology-biology.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Homology is the relationship between biological features that derive from a corresponding feature in a common ancestor. The features may be anatomical structures, developmental processes, genes, proteins, or other identifiable components of organisms. A human arm, a bat wing, and a whale flipper are homologous as vertebrate forelimbs, despite their different appearances and functions. Conversely, features that perform similar tasks need not be homologous: the wings of birds and insects evolved independently as flight organs.\n\nThe concept makes biological comparison historical rather than merely descriptive. It allows differences among organisms to be interpreted as transformations of inherited structures and provides a foundation for evolutionary biology, phylogenetics, and comparative genomics. Establishing homology requires evidence about correspondence and ancestry; superficial resemblance alone is insufficient. The relationship must also be specified at an appropriate level, because a complex organ may contain ancient homologous components while possessing an independently evolved organization or function.\n\nIn modern biology, homology links evidence from fossils, anatomy, embryology, molecular sequences, and gene regulation. Its application ranges from reconstructing evolutionary relationships to identifying gene families and comparing experimental organisms. These uses share a historical principle, but they require different methods and encounter different uncertainties."
    },
    {
      "heading": "Historical development of the concept",
      "text": "Comparative anatomy recognized recurring structural arrangements before an evolutionary explanation of them became widely accepted. Vertebrate limbs were particularly influential: differently shaped appendages could be described using corresponding bones and articulations. Early nineteenth-century anatomists disagreed about how far such structural unity extended and whether it reflected common organizational plans, functional constraints, or transformations of a shared pattern.\n\nRichard Owen gave the distinction between homology and analogy its influential nineteenth-century formulation. In 1843 he defined a homologue as “the same organ in different animals under every variety of form and function.” Analogy, by contrast, concerned functional correspondence. Owen interpreted structural relationships through an archetype rather than through the modern concept of genealogical descent. His usage therefore established an important comparative distinction without yet giving it its present evolutionary explanation. Étienne Geoffroy Saint-Hilaire had also emphasized the relative positions and connections of anatomical parts as evidence of correspondence.\n\nCharles Darwin supplied a historical explanation in On the Origin of Species in 1859. Descent with modification accounted for why related organisms retained the same underlying arrangement despite adapting to different conditions. The forelimb skeleton could persist because descendants inherited it, while natural selection modified its proportions and uses. Vestigial structures and similarities among embryos became further evidence of inherited organization rather than imperfect realizations of a fixed ideal. Homology consequently became a relationship of common origin, not an assertion that organs must remain similar in function or appearance.\n\nDuring the twentieth century, the development of genetics and molecular biology extended the concept to hereditary material. Comparisons of hemoglobin and other proteins showed how molecular differences could be interpreted as accumulated evolutionary changes. DNA sequencing later made it possible to compare hereditary information directly. In 1970 Walter Fitch introduced the distinction between orthologous and paralogous genes, separating common ancestry through species divergence from common ancestry through gene duplication. This terminology became essential as sequence databases expanded.\n\nThe rise of cladistics, especially through Willi Hennig’s work in the mid-twentieth century, also changed how homology was evaluated. Proposed correspondences could be tested against the distribution of characters on evolutionary trees. Meanwhile, developmental genetics revealed that organisms with markedly different anatomies often use related regulatory genes. These developments did not replace anatomical homology with molecular similarity. Instead, they made explicit that the histories of genes, developmental mechanisms, and complete structures must be investigated separately and then related to one another."
    },
    {
      "heading": "Anatomical correspondence and its evidence",
      "text": "An anatomical homology hypothesis identifies a particular structure in one organism with a corresponding structure in another. The comparison concerns historical continuity through ancestral populations, not physical identity between individuals. It normally draws on several kinds of evidence: position relative to neighboring structures, internal organization, connections to other tissues, development, and intermediate conditions in living or fossil organisms. No single criterion is universally decisive, and strong conclusions usually rest on agreement among independent observations.\n\nTetrapod forelimbs illustrate the importance of relational organization. A human arm, bat wing, and whale flipper contain a humerus, radius and ulna, wrist elements, and digits arranged in corresponding sequences. The bones differ greatly in proportions: bats have elongated digits supporting a flight membrane, whereas many whales possess shortened proximal limb elements and additional phalanges within certain digits. These differences do not erase the underlying correspondences. The evidence lies in the complete anatomical arrangement, development, and evolutionary context, rather than in a requirement that individual bones have identical shapes.\n\nFossils can connect structures whose correspondence would otherwise be difficult to recognize. The mammalian middle ear provides a well-supported example. The malleus and incus correspond respectively to the articular and quadrate bones involved in the jaw joint of earlier synapsids and other nonmammalian amniotes. Comparative anatomy, embryology, and a succession of fossil jaw and ear configurations support this interpretation. During mammalian evolution, the dentary–squamosal articulation became the principal jaw joint, while elements associated with the older articulation became incorporated into the sound-transmitting apparatus. Homology here explains a substantial change in both position and function.\n\nPaleontology also shows why gaps in living diversity can mislead comparisons. A structure may look unique when its intermediate relatives are extinct. Conversely, similar environments can produce striking resemblances among distantly related organisms. Anatomical evidence is therefore assessed together with a broader account of relationships, rather than by matching isolated shapes. Soft tissues and embryological observations can be especially important when fossil preservation captures only mineralized parts.\n\nHomology must be stated at a sufficiently precise anatomical level. Bird and bat forelimbs are homologous as tetrapod forelimbs, but their specialized flight surfaces originated independently. Bird flight feathers and the skin membrane of a bat wing are not corresponding inherited flight structures. A comparison described simply as “wing homology” therefore conceals several distinct questions. Identifying the feature, its boundaries, and the proposed ancestral condition is an essential part of the hypothesis, not merely a preliminary naming exercise."
    },
    {
      "heading": "Homology, analogy, and phylogenetic characters",
      "text": "The distinction between homology and analogy separates shared history from shared performance. Analogy describes correspondence in function, such as the use of insect wings and bird wings for powered flight. Homoplasy is a related but more explicitly phylogenetic concept: it refers to similarity that does not result from inheritance of the relevant character state from the common ancestor under consideration. Homoplasy includes independently evolved similarities and evolutionary reversals. A resemblance can therefore be useful for understanding function while being misleading about relationships.\n\nConvergent evolution produces similar adaptations in separate lineages. Streamlined bodies in dolphins and many fishes reflect similar mechanical demands of movement through water, but dolphins remain mammals with a mammalian anatomical inheritance. Convergence need not be limited to overall appearance. It can affect biochemical pathways, sensory systems, or individual molecular positions. Similar selective pressures can favor comparable solutions, although developmental constraints and the range of available ancestral variation influence which solutions arise.\n\nPhylogenetic analysis requires a distinction between a character and its states. Vertebrate forelimbs may be treated as corresponding structures, while particular arrangements of their bones are coded as alternative states. A shared derived state, called a synapomorphy, supports a grouping when it originated in the group’s common ancestor. A shared ancestral state, or symplesiomorphy, may be homologous but does not uniquely support that narrower group. Thus, possession of a vertebral column helps characterize vertebrates but does not, by itself, establish a special relationship between bats and birds within vertebrates.\n\nAn initial proposal of correspondence is often called primary homology. It may be based on position, detailed resemblance, or development. Its compatibility with other characters is then assessed through phylogenetic analysis, producing what is sometimes called a secondary homology assessment. The distinction emphasizes that recognition and testing are different stages. A proposed similarity may prove consistent with a single origin, or its distribution may require independent origins, reversals, or an incorrect initial identification.\n\nThese assessments depend on models and evidence rather than on an assumption that evolution always minimizes change. Parsimony seeks explanations requiring relatively few character-state transformations, whereas likelihood and Bayesian methods evaluate data under explicit evolutionary models. In molecular datasets, differences in evolutionary rate, biased nucleotide composition, and repeated substitutions can obscure historical signal. In morphological datasets, correlated characters or ambiguous boundaries can overweight particular complexes. Additional taxa, fossils, and independent datasets can change the interpretation of a character. Homology remains a historical claim, while support for that claim can strengthen or weaken as the inferred tree and the available observations change."
    },
    {
      "heading": "Molecular homology and sequence comparison",
      "text": "At the molecular level, homology concerns descent of nucleotide or amino-acid sequences from an ancestral sequence. DNA, RNA, and protein sequences change through substitutions, insertions, deletions, duplications, and rearrangements. Related sequences can therefore differ substantially while remaining homologous. Conversely, a short identical sequence need not demonstrate common ancestry: simple repeats, restricted chemical requirements, and chance matches can produce resemblance without providing persuasive evidence of a shared origin for the segment being compared.\n\nSequence alignment proposes correspondences among residues. In a pairwise alignment, matching positions are arranged in columns, with gaps representing inferred insertions or deletions. Multiple alignments extend this comparison to three or more sequences. An alignment is not a direct observation of evolutionary history; it is a hypothesis about which present-day residues descend from corresponding ancestral positions. Regions with little remaining signal may admit several plausible alignments, affecting downstream estimates of evolutionary relationships.\n\nSequence identity is a measurement, whereas homology is an ancestry relationship. For a specified alignment, percentage identity can be written as\n\n\\[I = 100 \\times \\frac{N_{\\mathrm{identical}}}{N_{\\mathrm{compared}}},\\]\n\nwhere the denominator must be defined, particularly when gaps or partial alignments are present. Two proteins may have 70 percent identity over a stated region; describing them as “70 percent homologous” confuses a quantitative resemblance with a historical relationship. The uncertainty lies in the evidence for homology, not in a percentage of ancestral relatedness. Similarity scores can additionally reward chemically conservative substitutions between different amino acids.\n\nDatabase-search programs such as BLAST, first described in 1990, identify local sequence similarities efficiently. Their scores and expectation values help assess whether a match is unlikely to arise in a specified background model. An expectation value estimates the number of matches of comparable or greater score expected by chance in a search of that scale. It is not the probability that two sequences are unrelated, and it does not by itself establish orthology or shared biological function. Sequence length, compositional bias, database size, and the extent of the alignment all matter.\n\nFor remote relationships, profile methods can detect patterns that a simple pairwise comparison misses. A family-wide profile represents conservation and variation across multiple positions, while profile hidden Markov models also describe insertions and deletions. Conserved three-dimensional structure may provide further evidence because structural constraints can preserve a fold after much sequence similarity has disappeared. Nevertheless, a short motif or common fold may be insufficient on its own. Molecular homology is most persuasive when statistically supported sequence correspondence agrees with domain organization, structural details, and a plausible evolutionary history."
    },
    {
      "heading": "Orthology, paralogy, and gene-family evolution",
      "text": "Homologous genes can be related through different historical events. Orthology describes genes whose relevant divergence traces to a speciation event. Paralogy describes genes whose relevant divergence traces to gene duplication. Both are forms of homology, but they answer different questions. A gene in one species can have several homologues in another species, and the closest sequence match is not necessarily the orthologue. Determining the relationship requires reconstructing the branching history of the gene family.\n\nConsider a single ancestral gene that duplicates into copies A and B before two species separate. The A copy in one descendant species is orthologous to the A copy in the other, and the two B copies are likewise orthologous. An A copy and a B copy are paralogous because their separation began with duplication. Paralogs can occur within the same species or in different species. Their classification depends on the ancestral event, not on whether the present-day genes happen to reside in one genome.\n\nDuplication can also occur after speciation. If one lineage retains a single gene while the other duplicates it, both resulting copies may be co-orthologous to the unduplicated gene in the first lineage. This produces a one-to-many relationship rather than a simple one-to-one correspondence. Further rounds of duplication and loss generate many-to-many relationships. If different descendant lineages lose different copies from an ancient duplication, their surviving single-copy genes can appear to be straightforward counterparts even though they are paralogs. Such differential loss complicates large-scale orthologue identification.\n\nThe vertebrate globin family illustrates the connection between duplication and functional diversification. Alpha-like and beta-like globins descend from an ancient duplication, and later duplications produced additional family members with distinct expression patterns. In humans, different globin genes contribute to hemoglobins predominant at embryonic, fetal, and adult stages. Common ancestry explains related sequences and folds, while changes in regulation and biochemical properties explain why the copies are not interchangeable in every context. Other duplicated genes can lose function or partition aspects of the ancestral role rather than acquiring a wholly new activity.\n\nHorizontal gene transfer adds another kind of branching history by moving genetic material between lineages outside ordinary parent-to-offspring descent. The term xenology is sometimes used for homologues related through such transfer, although terminology and operational definitions vary. In practice, gene trees are compared with species trees to infer duplications, losses, and transfers. These reconciliations can remain uncertain when trees are weakly resolved or taxon sampling is incomplete. Orthology is therefore valuable for functional comparison, but it is not a guarantee of identical function: even genes separated by speciation alone can evolve different expression patterns, interaction partners, or biochemical activities."
    },
    {
      "heading": "Domains, chromosomes, and genomic mosaics",
      "text": "Homology need not extend across an entire molecule. Many proteins contain multiple domains, each with its own evolutionary history. Duplication, fusion, fission, and recombination can place an old domain in a new molecular context. Two proteins may therefore share one homologous domain while possessing unrelated regions elsewhere. Whole-protein comparisons that ignore this modular organization can incorrectly imply that every segment has the same ancestry or that a local match establishes equivalent functions for both complete proteins.\n\nProtein kinases provide a clear example of recurring domains in different architectures. A conserved catalytic domain occurs in proteins that differ in regulatory segments, localization signals, and associated interaction domains. The common catalytic machinery supports a relationship among those domains, but the surrounding architecture influences how each protein is controlled and which substrates it encounters. Similar considerations apply to domains involved in DNA binding, membrane association, and protein–protein interactions. Domain-level annotation can thus be more informative than assigning a single label to an entire sequence.\n\nAt the genomic scale, synteny and conserved gene order provide evidence independent of sequence resemblance alone. In modern comparative genomics, conserved synteny commonly refers to genomic regions containing corresponding sets of genes, sometimes with substantial rearrangement of their order. Closely related species may retain long collinear blocks, whereas inversions, translocations, duplication, and deletion progressively alter them. If a rapidly evolving gene lies between well-established corresponding neighbors in two genomes, that positional information can help identify its counterpart when sequence comparison is inconclusive.\n\nHomologous chromosomes in a diploid organism are corresponding parental chromosome copies carrying largely the same set of genes at corresponding loci. They need not have identical sequences, because they can contain different alleles and structural variants. Their pairing during meiosis facilitates segregation and crossing over. Sex chromosomes illustrate the limits of simple chromosome-wide correspondence: in many mammals, X and Y chromosomes retain pairing and recombination in restricted pseudoautosomal regions while differing extensively elsewhere. Chromosome homology therefore does not imply uniform similarity along their entire lengths.\n\nHomologous recombination is named for its use of corresponding DNA sequences during strand exchange and repair. It can operate between homologous chromosomes, sister chromatids, or other sufficiently similar DNA copies. Its immediate molecular requirements are related to sequence matching and enzymatic recognition, whereas evolutionary homology concerns historical origin. Meanwhile, endosymbiotic theory explains why nuclear and organellar genes can have different deeper ancestries: mitochondria and chloroplasts descend from bacterial endosymbionts, and many genes were subsequently transferred to the nucleus. A genome is consequently a collection of overlapping histories, not a single indivisible hereditary narrative."
    },
    {
      "heading": "Developmental mechanisms and deep homology",
      "text": "Evolutionary developmental biology investigates how changes in developmental systems generate differences among organisms. It has shown that conserved regulatory components can participate in the formation of structures with very different anatomies. This finding broadens the evidence relevant to homology but also demands caution: sharing a developmental gene does not automatically establish that the complete structures using it descend from the same ancestral organ.\n\nThe Hox genes are an important example. These genes encode transcription factors involved in specifying positional identities along the anterior–posterior axis of many animals. Their related DNA-binding homeodomains and evolutionary histories establish molecular homology. In several animal groups, their arrangement in genomic clusters is associated with patterns of expression along the body axis, although cluster organization and regulatory relationships vary. Corresponding Hox genes can influence different morphological outcomes in different lineages because they act within broader networks of targets, signals, and tissue interactions.\n\nDeep homology describes situations in which ancient, homologous genetic or developmental machinery underlies structures whose detailed anatomical relationship is more distant or whose specialized forms evolved independently. The involvement of Pax6-family genes in eye development helped make this concept influential. In 1995, experiments showed that expression of the Drosophila eyeless gene could induce ectopic eye structures, and a mouse Pax6 gene could activate eye development in the fly experimental system. The structures produced reflected the developmental capabilities of fly tissues; the experiments did not convert flies into organisms bearing mammalian eyes.\n\nThese results demonstrate conservation of regulatory capacity, not identity of entire developmental programs. Eyes include photoreceptors, optical structures, pigment cells, support tissues, and neural connections whose histories can differ. Shared photoreceptive components may be ancient even where image-forming arrangements arose separately. Reuse of an existing regulatory gene in a new context, often called co-option, provides a mechanism by which molecular homology and independently evolved organ-level similarity can coexist.\n\nDevelopmental mechanisms can also change while a recognizable homologous structure persists. This phenomenon is often termed developmental system drift. Compensating changes in regulatory interactions can preserve an anatomical outcome despite alterations in the underlying developmental network. Conversely, conserved genes can be deployed in new organs. Neither developmental sameness nor developmental difference is therefore a sufficient universal test of structural homology. Comparative embryology, gene expression, perturbation experiments, lineage relationships, and anatomy must be evaluated together, with explicit attention to which biological level is being compared."
    },
    {
      "heading": "Serial homology, plant organs, and repeated organization",
      "text": "Serial homology concerns corresponding structures repeated within an organism, such as vertebrae, ribs, or appendages. It differs from the familiar comparison of a structure between species, but both involve repeated use and modification of inherited organization. Serially corresponding parts can become highly specialized: cervical, thoracic, and lumbar vertebrae occupy different regions and perform different mechanical roles while retaining relationships to a repeated vertebral pattern.\n\nArthropod appendages illustrate extensive specialization of repeated units. Antennae, mouthparts, walking legs, and other appendages differ in shape and function, but comparative anatomy and development reveal correspondences among their components and segmental positions. Regional gene regulation helps assign different identities to appendages along the body. Such observations do not mean that every repeated organ is produced by a literally identical program. The extent of correspondence must be established, and appendage elements can be lost, fused, or transformed during evolution.\n\nVertebrate forelimbs and hindlimbs also share aspects of developmental organization, including related skeletal arrangements and many signaling pathways, while differing in positional identity and characteristic regulatory inputs. Their comparison illustrates why serial homology and between-species homology must be distinguished. A human arm corresponds historically to another tetrapod’s forelimb; its relation to the human leg concerns a different axis of comparison. Similar molecular components contribute to both, but the evolutionary origin of paired appendages and the differentiation of anterior and posterior appendages pose additional historical questions.\n\nPlants provide especially clear examples of organs that differ in appearance while retaining developmental and positional relationships. A cactus spine is a modified leaf, whereas many thorns, including those of hawthorn, are modified shoots. Rose prickles are superficial outgrowths rather than modified branches. These defensive structures can be analogous in function without being homologous as the same type of organ. Tendrils likewise have different origins in different plant groups: pea tendrils derive from leaf parts, while grapevine tendrils belong to the shoot system. Their climbing function alone does not identify their ancestry.\n\nFloral organs have long been interpreted through comparisons with vegetative organs and with one another, notably in Goethe’s 1790 work on plant metamorphosis. Modern developmental genetics supplies mechanisms for the specification of sepals, petals, stamens, and carpels, including interactions among floral identity genes. Yet a conserved gene family is not a complete account of each organ’s evolutionary origin. Plants repeatedly produce organs from meristems, and similar positions or forms may arise through different developmental modifications. Anatomical continuity, vascular organization, developmental origin, fossil evidence, and phylogenetic distribution together help distinguish inherited correspondence from repeated functional innovation."
    },
    {
      "heading": "Uses in comparative biology and biotechnology",
      "text": "Homology enables observations from one organism to inform research on another. If corresponding genes or cellular components retain related roles, experimentally tractable organisms can reveal mechanisms that would be harder to investigate directly elsewhere. Studies of yeast cell-cycle control, for example, identified regulatory systems with homologues in other eukaryotes. The inferential strength comes from combined molecular, genetic, and functional evidence, not simply from the fact that two sequences resemble each other.\n\nGenome annotation relies heavily on such comparisons. When a newly sequenced genome contains a sequence related to a characterized enzyme, conserved catalytic residues, domain organization, genomic context, and phylogenetic placement can support a proposed molecular role. More specific assignments require more evidence. Members of the same homologous enzyme family may act on different substrates, and a conserved catalytic fold does not establish an identical physiological function. Automated transfer of an overly precise annotation can propagate errors through databases if later assignments reuse earlier predictions as though they were independent experimental findings.\n\nComparative genomics also uses homology to identify constrained sequences. A DNA region that remains unusually conserved across appropriately chosen species may be subject to selection preserving a molecular function. Such regions can include protein-coding exons, structural RNAs, and regulatory elements. Conservation does not identify the function by itself, and lack of readily detectable conservation does not demonstrate lack of function. Recently evolved elements, rapid sequence turnover, and compensatory changes can all limit straightforward cross-species comparisons.\n\nStructural modeling has historically benefited from homologous templates. If the structure of one protein is known, an aligned relative can often be modeled using the shared fold, with uncertainty concentrated in variable loops, insertions, and poorly conserved regions. Newer structure-prediction approaches have greatly expanded the available methods, but evolutionary comparisons remain important for identifying conserved sites, interpreting families, and distinguishing plausible structural correspondences. Structural similarity and ancestry must still be separated from claims about binding specificity or cellular activity.\n\nHomology is also central to practical biotechnology. Homologous recombination can direct DNA integration using flanking sequences corresponding to a target locus, and comparative analyses help identify candidate genes for experimental study. In medical genetics, evidence from homologous proteins, conserved residues, and model-organism experiments can contribute to interpreting variants. Such evidence is contextual rather than conclusive: biochemical effects, gene dosage, tissue expression, and organism-specific biology can differ. A well-supported evolutionary relationship establishes a basis for comparison, not automatic equivalence between organisms or a standalone clinical interpretation."
    },
    {
      "heading": "Limits of inference and unresolved questions",
      "text": "The historical definition of homology is clear, but many actual comparisons remain uncertain. Ancient divergence can erase detectable sequence similarity, fossils preserve only selected aspects of organisms, and developmental mechanisms can be reorganized. Absence of recognizable correspondence is therefore not equivalent to evidence of independent origin. At the same time, invoking common ancestry without distinguishing among possible histories can make a claim too broad to test. Useful hypotheses specify the feature, its boundaries, the organisms compared, and the ancestral relationship proposed.\n\nA particularly important limitation is that universal common ancestry does not make every feature homologous to every other feature. Biological structures arise, duplicate, combine, and disappear within lineages. Two proteins can share an ancient domain without sharing the ancestry of their entire architectures. Two organs can use homologous signaling molecules without descending from the same ancestral organ. The broad relatedness of organisms supplies the framework for these questions, but a specific homology claim concerns continuity of a specified character or component.\n\nGene histories can also disagree with species histories. Duplication, loss, and horizontal transfer are major causes, but incomplete lineage sorting can produce discordant gene trees even without those events. If ancestral genetic variants persist through closely spaced speciation events, their later coalescence can follow a different branching order from the species separations. A gene-tree disagreement therefore does not by itself identify an erroneous homology assignment or prove transfer. Distinguishing explanations requires suitable sampling, explicit models, and information from multiple genomic regions.\n\nComplex traits expose another difficulty: their boundaries are often partly analytical choices. An eye, immune pathway, or cell type comprises many interacting components that need not have originated together. Comparative single-cell studies now assess cell-type relationships using patterns of gene expression, regulatory activity, anatomy, and developmental origin. A similar expression profile can reflect shared ancestry, similar physiological demands, or both. Establishing cell-type homology consequently requires distinguishing stable lineage-specific organization from temporary cell states and from convergent expression of common functional machinery.\n\nFinally, no fixed percentage of sequence identity or degree of anatomical resemblance provides a universal threshold for homology. Detectability depends on sequence length, rates of evolution, structural constraints, background composition, and the availability of intermediate comparisons. Highly conserved short motifs may be uninformative, whereas a long, weak but coherent alignment can support remote ancestry. In difficult cases, alternative interpretations remain explicitly testable through new taxa, improved genomes, fossils, structures, and developmental experiments. The unresolved issue is often not whether organisms share ancestry, but exactly which parts correspond and how those parts were transformed."
    }
  ],
  "links": [
    "gene",
    "protein",
    "evolution",
    "phylogenetics",
    "charles-darwin",
    "natural-selection",
    "genetics",
    "molecular-biology",
    "hemoglobin",
    "paleontology",
    "water",
    "dna",
    "rna",
    "amino-acid",
    "orthology",
    "horizontal-gene-transfer",
    "protein-domain",
    "chromosome",
    "allele",
    "meiosis",
    "chromosomal-crossover",
    "homologous-recombination",
    "endosymbiotic-theory",
    "transcription-factor",
    "cell-cycle",
    "enzyme"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}