{
  "id": "whole-exome-sequencing",
  "title": "Whole-Exome Sequencing",
  "lang": "en",
  "source_lang": "en",
  "summary": "Whole-exome sequencing analyzes primarily protein-coding regions of the genome to identify genetic variants for research and clinical investigation.",
  "category": "science",
  "subcategory": "biology",
  "type": "concept",
  "aliases": [
    "Exome sequencing",
    "WES"
  ],
  "generated_at": "2026-11-03",
  "model": "pro/gpt-6.1-sol",
  "url": "https://aiwiki.page/en/whole-exome-sequencing/",
  "translations": {
    "zh-Hans": {
      "url": "https://aiwiki.page/zh/whole-exome-sequencing/",
      "json": "https://aiwiki.page/zh/whole-exome-sequencing.json"
    }
  },
  "sections": [
    {
      "heading": "",
      "text": "Whole-exome sequencing (WES) is a DNA sequencing approach designed to analyze most of the protein-coding regions of a genome in a single experiment. These regions constitute the principal target of the exome, which accounts for roughly 1–2% of the human genome, depending on its definition. By concentrating sequencing effort on regions with well-established biological functions, WES can identify variants relevant to inherited disorders, cancer, and other research questions while generating substantially less sequence data than whole-genome sequencing. Despite its name, it does not provide complete or equally reliable coverage of every targeted region."
    },
    {
      "heading": "Biological basis and scope",
      "text": "In many human genes, sequences retained in mature RNA are interrupted in genomic DNA by introns. During RNA splicing, introns are removed and exons are joined. Protein-coding portions of the resulting messenger RNA specify the sequence of a protein.\n\nAn exon is not necessarily entirely protein-coding: exons can also contain untranslated regions, and noncoding RNA genes have exons. In practice, conventional WES assays primarily target annotated protein-coding exons and often include short stretches of adjacent intronic sequence, where variants can disrupt splicing. Some designs additionally capture selected untranslated regions or other medically relevant sequences. The regions analyzed therefore depend on the capture kit, gene annotation, and laboratory pipeline.\n\nThe emphasis on coding sequence reflects its interpretability. A variant that changes a protein sequence, introduces a premature stop signal, or disrupts a splice site may have a more readily identifiable consequence than a variant in a poorly characterized regulatory region. Nevertheless, many coding variants are harmless, while important disease-causing variants occur outside the exome."
    },
    {
      "heading": "Laboratory and computational workflow",
      "text": "Sample preparation and target enrichment\n\nDNA is commonly extracted from blood, saliva, or tissue. Its quality and quantity influence subsequent performance. The DNA is fragmented, and sequencing adapters are attached to create a library.\n\nMost WES protocols enrich this library by hybridization capture. Synthetic probes complementary to selected genomic regions bind matching DNA fragments. The bound fragments are isolated and sequenced, while much of the untargeted DNA is removed. Amplification using the polymerase chain reaction may be included during library preparation.\n\nCapture is not perfectly selective or uniform. Some targeted regions yield many reads, others yield few, and some untargeted fragments are also sequenced.\n\nSequencing and alignment\n\nEnriched libraries are generally analyzed using high-throughput, short-read sequencing. Computational tools align the resulting reads to a reference genome and assess mapping quality, sequencing errors, contamination, and duplicated fragments.\n\nImportant quality measures include:\n\n- Depth of coverage: the number of reads covering a position.\n- Breadth of coverage: the proportion of target positions meeting a specified coverage threshold.\n- Mapping quality: confidence that a read has been assigned to the correct genomic location.\n- Coverage uniformity: how evenly sequencing reads are distributed across targeted regions.\n\nA high average depth does not guarantee that every clinically relevant exon has adequate coverage.\n\nVariant detection and annotation\n\nVariant-calling software identifies differences between the sample and the reference sequence. WES is particularly suited to detecting single-nucleotide variants and small insertions or deletions in adequately covered regions.\n\nSome pipelines also infer copy-number variants, such as deletions or duplications involving one or more exons, from changes in read depth. Performance varies by assay, event size, and genomic context; this capability requires separate validation rather than being an automatic property of exome sequencing.\n\nDetected variants are annotated with information about affected genes, predicted sequence consequences, population frequency, inheritance patterns, and previously reported clinical associations."
    },
    {
      "heading": "Interpretation",
      "text": "Identifying a sequence difference is distinct from establishing that it explains a person's condition. Interpretation combines molecular evidence with the individual's phenotype, family history, and the known relationship between a gene and a disorder.\n\nFor inherited-disease analysis, relevant considerations include:\n\n- Whether the variant is compatible with the expected inheritance pattern.\n- Whether its population frequency is consistent with the disorder's prevalence and inheritance.\n- Whether its predicted effect matches an established disease mechanism.\n- Whether it segregates with the condition in a family.\n- Whether experimental or clinical evidence supports its significance.\n\nClinical laboratories commonly classify variants as pathogenic, likely pathogenic, of uncertain significance, likely benign, or benign. A variant of uncertain significance indicates insufficient or conflicting evidence; it is not equivalent to an established molecular diagnosis. Computational predictions alone generally do not establish pathogenicity.\n\nFamily-based sequencing\n\nA common design is trio sequencing, in which an affected individual and both biological parents are analyzed. This can help distinguish inherited variants from newly arising variants and determine whether two variants in a recessive-disease gene are likely to affect different parental copies.\n\nFamily data can substantially reduce the number of candidate explanations, but interpretation may still be complicated by incomplete penetrance, variable expression, mosaicism, or incomplete knowledge of gene–disease relationships."
    },
    {
      "heading": "Applications",
      "text": "Rare inherited disorders\n\nWES is used in the investigation of genetically heterogeneous conditions: disorders in which similar clinical features can result from variants in many different genes. Examples include some developmental, neurological, metabolic, and congenital disorders.\n\nIts diagnostic yield is not a fixed property of the technology. It depends on patient selection, prior testing, family structure, clinical information, assay quality, and current scientific knowledge. A negative result does not exclude a genetic cause.\n\nWES also supports the discovery of previously unrecognized disease-associated genes. Such discoveries typically require evidence beyond the observation of a rare variant in one individual.\n\nCancer research and molecular characterization\n\nIn cancer research, exome sequencing can characterize coding variants acquired by a tumor. Comparing tumor DNA with matched non-tumor DNA helps distinguish acquired changes from inherited variants.\n\nInterpretation is affected by tumor purity, genetic diversity among tumor cells, copy-number changes, and sequencing depth. Clinical cancer testing often uses targeted gene panels instead, because these can provide deeper coverage of selected genes or detect specific alterations more efficiently.\n\nPopulation and functional research\n\nLarge exome datasets support studies of rare coding variation, associations between genes and traits, and the effects of variants that disrupt gene function. These studies complement genome-wide association studies, which often examine common variants distributed throughout the genome.\n\nExome sequencing measures DNA sequence, not gene expression. It therefore cannot directly establish whether a gene is expressed or how its RNA is spliced in a particular tissue; RNA sequencing can provide complementary evidence."
    },
    {
      "heading": "Comparison with other sequencing approaches",
      "text": "Approach | Principal scope | Main distinction from WES\nTargeted gene panel | A selected set of genes or regions | Can offer deeper or more thoroughly validated coverage for a defined question, but has a narrower scope\nWhole-exome sequencing | Most annotated protein-coding regions, with assay-dependent additions | Broad coding-region analysis with relatively limited sequencing requirements\n[[whole-genome-sequencing | Whole-genome sequencing]] | Coding and noncoding genomic DNA\nRNA sequencing | RNA molecules present in a sample | Examines expression and transcript structure rather than genomic sequence directly\n\nA “virtual panel” may be applied to exome data by restricting initial analysis to selected genes. This limits interpretation rather than changing the regions originally captured and sequenced."
    },
    {
      "heading": "Technical and interpretive limitations",
      "text": "Incomplete capture and difficult genomic regions. Regions with extreme base composition, repetitive sequence, or close similarity to other genes can be poorly covered or incorrectly mapped. Some medically important genes contain such regions.\n\nLimited noncoding analysis. Most deep intronic and regulatory variants lie outside standard exome targets. Capturing short exon boundaries does not provide comprehensive coverage of sequences involved in regulation or splicing.\n\nIncomplete detection of complex variation. Repeat expansions, balanced rearrangements, many structural variants, and some exon-level deletions or duplications are not reliably detected by conventional WES. Mitochondrial DNA analysis also depends on the assay and cannot be assumed to be comprehensive.\n\nSensitivity to mosaicism and sample choice. A variant present in only a small proportion of cells may be missed. A variant confined to a particular tissue may not be present in the sampled material.\n\nKnowledge-dependent interpretation. Disease-gene catalogs, population datasets, and classification evidence change over time. Reanalysis can produce new findings without generating new sequence data, although it cannot recover regions that were never adequately sequenced.\n\nUneven population representation. Underrepresentation of some ancestry groups in reference datasets can make rare variants harder to interpret and contribute to differences in rates of uncertain findings.\n\nLaboratories may use complementary methods to investigate findings that are technically uncertain or outside the assay's validated capabilities."
    },
    {
      "heading": "Development and ethical considerations",
      "text": "WES developed from advances in massively parallel sequencing and targeted DNA enrichment during the late 2000s. It became an important bridge between sequencing individual genes and analyzing entire genomes, particularly for discovering causes of rare inherited disorders. Its relative advantages continue to depend on sequencing costs, laboratory infrastructure, analytical requirements, and the clinical or research question.\n\nBecause exome data cover many genes, analysis can reveal findings unrelated to the original reason for testing. Secondary findings generally refer to deliberately sought results outside that original question, whereas incidental findings arise unexpectedly; terminology and reporting practices vary.\n\nConsent and reporting policies address which findings are analyzed or returned, testing involving children, storage and reuse of data, and information relevant to biological relatives. Genomic data also raise data privacy concerns because they are potentially identifiable and contain information shared with family members."
    }
  ],
  "links": [
    "dna-sequencing",
    "genome",
    "gene",
    "dna",
    "rna-splicing",
    "messenger-rna",
    "protein",
    "exon",
    "polymerase-chain-reaction",
    "copy-number-variation",
    "phenotype",
    "mosaicism",
    "cancer",
    "genome-wide-association-study",
    "gene-expression",
    "rna-sequencing",
    "whole-genome-sequencing",
    "data-privacy"
  ],
  "sources": [],
  "license": "https://creativecommons.org/licenses/by/4.0/"
}