Whole-Exome Sequencing in Bioinformatics

  • Whole-exome sequencing (WES) is a DNA sequencing approach that focuses primarily on the protein-coding regions of an organism’s genome, known collectively as the exome. These regions contain the exons of protein-coding genes, which are transcribed into RNA and generally contribute to the production of proteins. Although protein-coding sequences represent only approximately 1–2% of the human genome, they contain a substantial proportion of known disease-associated genetic variants. Whole-exome sequencing is therefore widely used in medical genetics, rare disease research, cancer genomics, population genetics, and functional genomics. It provides a cost-effective alternative to whole-genome sequencing when the primary research objective is to investigate genetic variation in protein-coding regions.
  • The central principle of whole-exome sequencing is to selectively enrich DNA fragments corresponding to coding regions before sequencing. Instead of sequencing the entire genome, WES uses specialized capture methods to isolate or enrich exon-containing DNA fragments. The enriched DNA is then sequenced using next-generation sequencing technologies, most commonly short-read sequencing platforms. The resulting reads are analysed using bioinformatics pipelines to identify genetic variants, including single-nucleotide variants, small insertions and deletions, and selected other genomic alterations.
  • The exome consists primarily of protein-coding exons, but the exact regions included in a whole-exome sequencing experiment depend on the capture kit and its design. Many commercial exome capture systems also include selected untranslated regions, known disease-associated noncoding regions, splice-site regions, or other targeted sequences. Therefore, the term whole-exome sequencing does not mean that every exon or every clinically relevant genomic position is guaranteed to be captured. Coverage varies among genes and genomic regions, and some exons may be poorly represented or absent from a particular capture design.
  • Whole-exome sequencing differs from whole-genome sequencing (WGS) and targeted gene sequencing. WGS aims to analyse genomic regions across the entire genome, including coding and noncoding sequences. WES concentrates on protein-coding regions and selected additional targets, reducing the amount of DNA that needs to be sequenced. Targeted gene panels focus on a predefined group of genes or genomic regions associated with a particular disease or biological process. WES examines a much broader set of genes than most targeted panels, making it useful when the genetic cause of a condition is uncertain or when several different genes may be involved.
  • The workflow of whole-exome sequencing begins with the collection of a biological sample. In human genetic research, blood is commonly used, although saliva, buccal swabs, tissue, and other suitable materials may also provide genomic DNA. For cancer research, tumour tissue may be analysed, sometimes together with matched normal tissue. The sample must be correctly identified, collected, stored, and processed to minimize contamination and DNA degradation. The choice of sample depends on the research question, the biological material available, and the requirements of the sequencing protocol.
  • DNA extraction is performed to isolate genomic DNA from the sample. The extraction process generally involves cell lysis, removal of proteins and other contaminants, purification of nucleic acids, and recovery of DNA. DNA concentration is commonly measured using fluorometric methods such as Qubit, while absorbance-based measurements can provide additional information about purity. DNA integrity is also evaluated because highly degraded DNA may affect library preparation and capture efficiency. The quality requirements depend on the sample type and the exome sequencing protocol.
  • After extraction, DNA is prepared for sequencing. In a typical short-read WES workflow, genomic DNA is fragmented into smaller pieces using mechanical or enzymatic methods. The fragment size is selected according to the library preparation and capture system. Sequencing adapters are then attached to the DNA fragments. These adapters contain sequences required for amplification and sequencing. Sample-specific index sequences may also be added, allowing multiple samples to be pooled and sequenced together in a process known as multiplexing.
  • The prepared DNA library is subjected to target enrichment, which is the defining stage of whole-exome sequencing. Target enrichment is generally performed using hybridization-based capture or related probe-based methods. In hybridization capture, specially designed oligonucleotide probes bind to DNA fragments corresponding to the selected exonic regions. These probes are often attached to biotin, which allows the probe–DNA complexes to be recovered using streptavidin-coated magnetic beads. Uncaptured DNA fragments are removed through washing, leaving an enriched collection of DNA fragments representing the targeted exome regions.
  • Different commercial systems use different probe designs, capture chemistries, target definitions, and enrichment protocols. Examples include Agilent SureSelect, Roche NimbleGen SeqCap, and Twist exome capture systems. Some workflows use one round of hybridization capture, while others may use additional enrichment or amplification steps. The choice of capture system influences which exons are represented, the uniformity of coverage, and the ability to capture difficult genomic regions. Researchers should examine the target-region files and technical documentation supplied with the selected kit.
  • After target enrichment, the captured DNA fragments are amplified when required, purified, quantified, and assessed for library quality. The enriched library is then loaded onto a sequencing platform. Illumina sequencing is commonly used for WES because it provides high-throughput short reads with high per-base accuracy. Paired-end sequencing is frequently used, in which both ends of each DNA fragment are sequenced. This provides additional information about fragment orientation and insert size, which can improve read alignment and variant detection.
  • The amount of sequencing performed is described by sequencing depth or coverage. In whole-exome sequencing, coverage refers to the number of sequencing reads representing a targeted genomic position. Average target coverage is commonly reported as a multiple, such as 50×, 100×, or 200×. However, the average coverage alone does not describe how evenly the exome is represented. Some regions may have very high coverage, while others may have low coverage or no reads at all. Therefore, researchers commonly report the percentage of target bases covered at or above specified thresholds, such as 10×, 20×, or 30×.
  • Coverage uniformity is an important consideration in WES because capture efficiency varies across genomic regions. GC-rich sequences, repetitive regions, homologous genes, regions with complex sequence composition, and poorly designed capture probes may be difficult to enrich or sequence. Some exons may also be missed because of differences between the reference genome and the sample. Consequently, a high average sequencing depth does not guarantee that every clinically or biologically important exon has adequate coverage.
  • The theoretical sequencing requirement can be estimated using the following relationship:
  • Coverage = Total number of sequenced bases / Target region size
  • For example, if 100 million paired-end read pairs are generated and each read is 150 bases long, the total number of sequenced bases is:

Total bases = 100,000,000 × 2 × 150

            = 30,000,000,000 bases

  • If the effective target region is 50 million bases, the theoretical average target coverage would be:

Coverage = 30,000,000,000 / 50,000,000

         = 600×

  • This is an idealized calculation. Actual target coverage depends on the proportion of reads successfully mapped to the targeted regions, capture efficiency, duplicate reads, sequencing quality, and coverage distribution. The effective target size also depends on the capture design and analysis definition.
  • After sequencing, the instrument produces raw sequencing files, commonly in FASTQ format. Each FASTQ record contains a sequence identifier, the nucleotide sequence, a separator line, and quality scores for individual bases. The quality scores are generally represented using the Phred scale. Higher Phred scores indicate a lower estimated probability of an incorrect base call. FASTQ files are the starting point for most short-read WES bioinformatics workflows.
  • The first computational stage is quality control of the raw sequencing reads. Quality control evaluates base quality, read length, GC content, adapter contamination, duplication levels, and other sequencing characteristics. FastQC is widely used to generate quality-control reports for FASTQ files. MultiQC can combine reports from several samples into a single summary. fastp can perform adapter trimming, quality filtering, and other preprocessing operations. These tools help identify technical problems before reads are aligned to a reference genome.
  • A simplified quality-control workflow can be performed using the following commands:

# Examine raw paired-end FASTQ files

fastqc sample_R1.fastq.gz sample_R2.fastq.gz

# Combine quality-control reports

multiqc .

# Remove adapters and filter low-quality reads

fastp \

    -i sample_R1.fastq.gz \

    -I sample_R2.fastq.gz \

    -o sample_R1_clean.fastq.gz \

    -O sample_R2_clean.fastq.gz \

    –html fastp_report.html \

    –json fastp_report.json

  • Filtering parameters should be selected according to the sequencing platform, library preparation protocol, and analytical objective. Excessive trimming may remove useful sequence information, while insufficient filtering may allow technical artifacts to affect downstream analysis. Quality-control reports should therefore be reviewed before proceeding.
  • After quality control, reads are aligned to a reference genome. Alignment assigns each sequencing read to its most likely genomic location. The reference genome must be appropriate for the organism and analysis, and its assembly version should be recorded. For human WES, commonly used reference assemblies include GRCh37 and GRCh38. Different reference assemblies can produce differences in genomic coordinates, variant annotation, and interpretation, so researchers should avoid mixing reference versions within a single analysis.
  • BWA-MEM2 is widely used for aligning short DNA reads to reference genomes. Bowtie2 is another short-read alignment tool, although BWA-MEM2 is commonly used in germline exome workflows. The resulting alignment files are generally stored in SAM or BAM format. SAMtools can be used to convert, sort, index, and inspect these files. Alignment quality should be evaluated using mapping rate, mapping quality, insert-size distribution, duplicate levels, coverage depth, and the percentage of reads mapped to target regions.
  • A simplified alignment workflow using BWA-MEM2 and SAMtools is:

# Align paired-end reads to the reference genome

bwa-mem2 mem reference.fa \

    sample_R1_clean.fastq.gz \

    sample_R2_clean.fastq.gz \

    > sample.sam

# Convert SAM to BAM and sort the alignments

samtools view -b sample.sam | \

    samtools sort -o sample.sorted.bam

# Index the sorted BAM file

samtools index sample.sorted.bam

# Generate alignment statistics

samtools flagstat sample.sorted.bam

  • The reference genome must be indexed using the commands required by the selected aligner before mapping. Real analysis workflows may also include read-group information, duplicate marking, base-quality assessment, and other processing steps. Alignment parameters and reference files should be consistent across samples to support reliable comparisons.
  • After alignment, duplicate reads are often identified and marked. PCR amplification during library preparation can produce multiple sequencing reads derived from the same original DNA fragment. These duplicates may artificially increase apparent coverage and influence variant detection. Tools such as Picard MarkDuplicates and SAMtools can support duplicate identification and marking, depending on the workflow. Unique molecular identifiers, when incorporated into the library, can provide additional information for distinguishing original molecules from amplification duplicates. Duplicate handling should be appropriate to the library design and sequencing method.
  • Variant calling is one of the main objectives of whole-exome sequencing. It identifies genetic differences between the sample and the reference genome. The most commonly analysed variants are single-nucleotide variants (SNVs) and small insertions and deletions (indels). SNVs involve a change in one nucleotide, while indels involve the insertion or deletion of a small number of nucleotides. These variants may alter protein sequences, affect RNA splicing, modify gene regulation, or have no detectable functional effect.
  • GATK HaplotypeCaller is widely used for germline small-variant calling. It analyses aligned reads and performs local haplotype assembly around candidate variant regions. FreeBayes and bcftools provide alternative approaches for variant detection. DeepVariant uses machine-learning methods to identify variants from sequencing data and supports suitable short-read workflows. The choice of caller depends on the organism, sample type, sequencing platform, ploidy, and research objective.
  • A simplified example of variant calling using GATK HaplotypeCaller is:

# Generate a genomic variant call file

gatk HaplotypeCaller \

    -R reference.fa \

    -I sample.sorted.bam \

    -O sample.g.vcf.gz \

    -ERC GVCF

  • This command illustrates single-sample gVCF generation. In multi-sample germline analysis, individual gVCF files can be combined and jointly genotyped using appropriate GATK procedures. Additional steps are required for a complete analysis, including quality assessment, variant filtering, normalization where appropriate, and annotation. The recommended workflow should follow the documentation for the GATK version being used.
  • Variant filtering is performed to reduce false-positive calls and retain variants supported by reliable sequencing evidence. Filtering may consider read depth, mapping quality, base quality, allele balance, strand bias, variant quality, and other metrics. In population or case-control studies, additional filtering may consider allele frequency, inheritance patterns, and consistency across samples. Hard-filtering thresholds and statistical models should be selected according to the dataset rather than applied indiscriminately.
  • The resulting variant calls are commonly stored in Variant Call Format (VCF) files. A VCF file contains information about chromosome position, reference allele, alternative allele, quality, filtering status, and supporting annotations. Compressed VCF files commonly use the .vcf.gz extension and can be indexed for efficient access. Tools such as bcftools and tabix support filtering, indexing, and querying variant files.
  • Variant annotation is performed to determine the genomic location and potential biological consequences of detected variants. Annotation can establish whether a variant occurs in a coding exon, splice site, untranslated region, or another genomic feature. Coding variants may be classified as synonymous, missense, nonsense, or frameshift variants. A synonymous variant changes a codon without changing the encoded amino acid, while a missense variant changes one amino acid. Nonsense variants introduce a premature stop codon, and frameshift variants alter the reading frame through insertions or deletions that are not multiples of three nucleotides.
  • Several tools are used for variant annotation. Ensembl Variant Effect Predictor (VEP) identifies predicted effects on genes and transcripts. SnpEff annotates variants using genomic feature databases, while ANNOVAR supports functional annotation and the integration of several databases. Annotation results depend on the reference genome, transcript definitions, database versions, and selected annotation settings. Researchers should record these details because different annotation releases may produce different consequences for the same genomic variant.
  • A simplified VEP annotation command may look like this:

# Annotate variants using Ensembl VEP

vep \

    -i sample_variants.vcf.gz \

    -o sample_annotated.vcf \

    –cache \

    –assembly GRCh38 \

    –everything

  • The annotation cache and reference assembly must match the intended analysis. Additional options may be needed for population frequencies, clinical annotations, transcript selection, and output customization. The command is illustrative and should be adapted to the installed VEP version and available databases.
  • Whole-exome sequencing is particularly useful in the investigation of rare genetic diseases. Many rare disorders are caused by variants in protein-coding genes, making WES a valuable method for identifying potential disease-associated mutations. It is often used when a patient’s clinical features suggest a genetic condition but the responsible gene is unknown. By examining thousands of genes simultaneously, WES can identify candidate variants that may explain the observed phenotype.
  • In rare disease research, variant interpretation commonly involves filtering variants by population frequency, predicted functional effect, inheritance pattern, and phenotype compatibility. For example, a rare recessive condition may involve two disease-associated variants in the same gene, while a dominant condition may be associated with a single heterozygous variant. However, inheritance patterns vary among disorders, and interpretation requires careful consideration of family relationships, penetrance, variable expressivity, and other genetic factors.
  • Family-based or trio whole-exome sequencing involves analysing samples from an affected individual and both biological parents. This approach can help identify de novo variants, inherited variants, and compound heterozygous variants. Trio analysis can reduce the number of candidate variants and provide valuable information about inheritance. Additional family members may also be included when they help clarify segregation patterns or disease-associated variants.
  • Whole-exome sequencing is also used in cancer genomics. Tumour exome sequencing can identify coding mutations in genes involved in cell growth, DNA repair, signalling pathways, and other cancer-related processes. When tumour tissue is analysed alongside matched normal tissue, researchers can distinguish many acquired somatic variants from inherited germline variants. This is important because tumour-specific mutations and inherited susceptibility variants have different biological and clinical implications.
  • Cancer WES can support the identification of potentially actionable mutations, investigation of tumour heterogeneity, and characterization of molecular pathways involved in cancer development. However, WES may not detect all cancer-associated changes. Large structural variants, copy-number alterations, noncoding regulatory changes, and complex rearrangements may be missed or incompletely characterized. Specialized analysis methods and complementary sequencing approaches may therefore be necessary.
  • Whole-exome sequencing can also be used in population genetics and evolutionary research. It provides information about variation in protein-coding regions across individuals, populations, or species. Researchers can investigate allele frequencies, genetic diversity, population structure, selection signals, and relationships among populations. Because WES targets selected genomic regions, however, it does not provide the same genome-wide representation as whole-genome sequencing. Capture bias, differences in target designs, and the influence of natural selection on coding regions must be considered when interpreting population-level results.
  • In agriculture and animal breeding, WES can help identify variants associated with economically important traits, inherited disorders, fertility, growth, production, and disease resistance. It can support the discovery of candidate genes and coding variants in livestock, poultry, companion animals, and crop species. In crop research, exome sequencing can investigate genes associated with yield, quality, stress tolerance, and disease resistance. In animal breeding, it can contribute to genetic characterization and the investigation of variants that influence production or health. However, many complex traits are influenced by variants across coding and noncoding regions, so WES alone may not capture all relevant genetic information.
  • One important limitation of whole-exome sequencing is incomplete and uneven target coverage. Capture probes may fail to enrich some exons, particularly those with unusual GC content, repetitive sequences, high sequence similarity, or complex genomic organization. Some genes contain highly homologous regions or pseudogenes that complicate read alignment and variant detection. Consequently, a negative WES result does not exclude a genetic disorder or establish that no relevant variant exists. Poorly covered regions may require targeted resequencing or another testing method.
  • Another limitation is that WES primarily examines coding regions and selected additional targets. It may miss variants in distant regulatory elements, deep intronic regions, intergenic regions, and other noncoding sequences not included in the capture design. It may also have limited sensitivity for certain structural variants, repeat expansions, copy-number changes, mitochondrial variants, and complex genomic rearrangements. The ability to detect these variant classes depends on the capture kit, sequencing depth, library design, and analytical pipeline.
  • Whole-exome sequencing may also generate variants of uncertain significance (VUS). A VUS is a genetic variant for which available evidence is insufficient to establish whether it is benign or disease-causing. Such variants should not be treated as confirmed explanations for a disease. Interpretation may require population frequency data, functional studies, segregation analysis, clinical observations, and additional evidence. Databases such as ClinVar, gnomAD, ClinGen, and disease-specific resources can support interpretation, but their records must be evaluated according to evidence quality and relevance.
  • The interpretation of WES results may also be affected by reference bias, sequencing errors, mapping ambiguity, sample contamination, and differences in transcript annotation. A variant may appear to have different predicted consequences depending on the transcript selected. Furthermore, a variant that changes a protein sequence is not necessarily harmful. Functional interpretation requires evidence about the gene, transcript, biological mechanism, population frequency, inheritance pattern, and relevant phenotype.
  • For reproducible WES analysis, researchers should document the biological sample, DNA extraction procedure, library preparation method, capture kit, target-region definition, sequencing platform, read length, and sequencing depth. The reference genome assembly, software versions, annotation databases, and analysis parameters should also be recorded. Quality-control reports should include target coverage, coverage uniformity, percentage of target bases meeting selected depth thresholds, mapping rate, duplicate levels, and variant quality measures. Original FASTQ files, alignment files, variant files, and workflow scripts should be retained according to applicable data-management and privacy requirements.
  • Whole-exome sequencing offers several advantages. It focuses sequencing resources on protein-coding regions, which contain many known disease-associated variants. It is generally less expensive and produces less data than sequencing the entire genome at comparable coverage objectives. It can examine thousands of genes simultaneously and is useful when a disease may involve one of many possible genes. It also supports research in cancer, population genetics, evolutionary biology, and agriculture.
  • Despite its advantages, WES is not a complete substitute for whole-genome sequencing. WGS provides broader coverage of coding and noncoding regions, while WES offers a more focused and often more economical approach to investigating protein-coding variation. Targeted gene panels may be more appropriate when the suspected condition is associated with a well-defined group of genes and very high coverage is required. The selection of a sequencing method should therefore be based on the research question, expected variant types, sample characteristics, required coverage, and available resources.
  • Whole-exome sequencing is a major technology in modern bioinformatics because it provides a practical approach to analysing genetic variation in protein-coding regions. Its workflow combines DNA extraction, library preparation, exon capture, sequencing, quality control, reference alignment, variant calling, annotation, and biological interpretation. It has made important contributions to rare disease research, cancer genomics, population studies, and agricultural genetics. Understanding its principles, capture methods, analytical tools, advantages, and limitations is essential for designing reliable exome sequencing studies and interpreting their results accurately.
Author:

Leave a Reply

Your email address will not be published. Required fields are marked *