When planning a DNA sequencing study or choosing a genetic test, one of the first decisions is how much of the genome to sequence. The three most common options are Whole Genome Sequencing (WGS), Whole Exome Sequencing (WES) and targeted sequencing (gene panels).
Each approach has its own strengths, costs and limitations. In this guide, we compare WGS vs WES vs targeted sequencing side by side and explain when to use each — in research, clinical diagnostics and bioinformatics projects.
Quick overview
- WGS reads (almost) the entire genome — coding and non-coding regions.
- WES reads only the exome — the protein-coding exons, about 1–2% of the human genome.
- Targeted sequencing reads a selected set of genes or regions — from a handful to a few hundred genes.
Think of the genome as a library: WGS reads every book, WES reads every chapter that contains instructions for proteins, and targeted sequencing reads only the specific pages you already suspect are important.
What is Whole Genome Sequencing (WGS)?
WGS sequences the complete DNA of an organism — about 3.1 billion base pairs in humans — including exons, introns, promoters, enhancers, intergenic regions and the mitochondrial genome. No capture step is needed: DNA is fragmented and sequenced directly, which gives more uniform coverage than capture-based methods.
Typical depth: around 30× for human germline studies (higher for cancer).
Detects: SNVs, indels, copy number variants (CNVs), structural variants (SVs), repeat expansions, non-coding variants and mitochondrial variants — all in one test.
What is Whole Exome Sequencing (WES)?
WES uses probes to capture the exons of roughly 20,000 protein-coding genes before sequencing. Although the exome is only about 1–2% of the genome, it is where a large share of known disease-causing variants are found — which is why WES is so widely used for diagnosing genetic disorders.
Typical depth: around 100× or more, because capture efficiency varies between regions.
Detects: SNVs and small indels in coding regions very well; CNV detection is possible but less reliable; most structural variants and non-coding variants are missed.
New to exome sequencing? Read our beginner’s guide to Whole Exome Sequencing.
What is targeted sequencing?
Targeted sequencing focuses on a pre-defined set of genes or regions, selected by hybridisation capture or PCR amplicons. Examples include cancer hotspot panels, hereditary cancer panels (e.g. BRCA1/BRCA2 and related genes), cardiomyopathy panels and pharmacogenomics panels.
Typical depth: very high — often 500× to several thousand×, which allows detection of low-frequency variants such as somatic mutations in tumours or circulating tumour DNA (liquid biopsy).
WGS vs WES vs targeted sequencing: comparison table
| Feature | WGS | WES | Targeted panel |
|---|---|---|---|
| Region sequenced | Entire genome | Protein-coding exons (~1–2%) | Selected genes/regions |
| Typical depth | ~30× | ~100×+ | ~500× to 1000s× |
| Cost per sample | Highest | Medium | Lowest |
| Data size | Largest (often 100 GB+ raw data per human sample) | Medium (a few GB to ~10–20 GB) | Small |
| Coverage uniformity | Most uniform (no capture) | Variable (capture bias) | High within targets |
| SNVs / small indels | Yes (genome-wide) | Yes (coding regions) | Yes (targets only) |
| CNVs / structural variants | Best | Limited | Limited (panel-dependent) |
| Non-coding variants | Yes | Mostly no | Only if targeted |
| Low-frequency variants | Limited at 30× | Moderate | Best |
| Incidental findings | Most | Some | Fewest |
| Turnaround and analysis | Longest, most compute | Moderate | Fastest, simplest |
When to use WGS
- Rare and undiagnosed diseases — especially when WES or panels have not found the cause.
- Structural variants and CNVs — large deletions, duplications, inversions and translocations.
- Non-coding and regulatory variants — promoters, enhancers, deep intronic and splicing variants.
- Population genomics and reference projects — such as national genome initiatives.
- Microbial and pathogen genomics — bacterial and viral genomes are small, so WGS is the standard approach for outbreak tracking and antimicrobial resistance.
- De novo genome assembly of new species (often with long reads).
When to use WES
- First-line diagnosis of suspected Mendelian disorders, especially when the clinical picture doesn’t point to a specific gene.
- Trio analysis (child + both parents) to find de novo variants.
- Large research cohorts where WGS cost is too high but broad coding coverage is needed.
- Cancer research focused on coding mutations, often with matched tumour–normal samples.
- Re-analysis over time — exome data can be re-analysed as new disease genes are discovered.
When to use targeted sequencing
- Well-defined clinical questions — e.g. testing known genes for hereditary breast and ovarian cancer.
- Cancer diagnostics and therapy selection — detecting actionable mutations at high depth.
- Liquid biopsy — finding very low-frequency tumour variants in blood.
- High-throughput screening of many samples at low cost.
- Pharmacogenomics — checking genes that affect drug response.
- Validating findings from WGS or WES in larger cohorts.
How to choose: a simple decision guide
- Do you know which genes matter? Yes → targeted panel. Not sure → WES or WGS.
- Are you looking for low-frequency (somatic) variants? Yes → targeted panel or deep WES.
- Do you need structural variants, CNVs or non-coding variants? Yes → WGS.
- Is budget or computing capacity limited? → targeted panel or WES.
- Is it a rare disease with negative previous tests? → WGS (or WES re-analysis).
- Is it a bacterial or viral genome? → WGS.
How the bioinformatics analysis differs
The core steps are similar for all three — QC, alignment (e.g. BWA-MEM), duplicate marking, variant calling (e.g. GATK HaplotypeCaller or DeepVariant) and annotation (VEP, ANNOVAR, SnpEff). The differences are in the details:
- WGS: the largest files and longest compute times; usually needs HPC or cloud resources; extra callers for CNVs and SVs (e.g. Manta, Delly, CNVnator).
- WES: requires the capture target BED file for coverage calculations and variant calling; coverage QC per target is critical; exome-specific CNV tools are used.
- Targeted panels: very high depth needs careful handling of duplicates and, for amplicons, primer trimming; low-frequency callers and UMIs (unique molecular identifiers) are often used for somatic detection.
Frequently asked questions
Is WGS better than WES?
WGS detects more types of variants and gives more uniform coverage, but it costs more and produces much more data. WES is often the most practical first test for coding variants; WGS is preferred when you need structural or non-coding variants, or when other tests are negative.
Why does WES need higher depth than WGS?
Capture efficiency varies between exons, so some regions get fewer reads. Sequencing to around 100× ensures most target regions have enough coverage for reliable variant calling.
Can WES detect CNVs?
Partly. Specialised tools can detect some CNVs from exome data, but sensitivity is lower than WGS, especially for small CNVs and those in non-coding regions.
Which is cheapest?
Targeted panels are usually the cheapest per sample, followed by WES, then WGS — although WGS costs have fallen sharply in recent years.
Which approach is best for learning bioinformatics?
WES is a great starting point: datasets are manageable on modest hardware, and the workflow teaches all the core steps of variant analysis that also apply to WGS and panels.
Learn WES and WGS data analysis with Bio Help Learning
Want hands-on experience with real sequencing data? At Bio Help Learning, our practical training takes you from raw FASTQ files to annotated, interpreted variants:
- Whole Exome Sequencing (WES) Data Analysis — available as live training and recorded sessions. The next live batch starts on 10 November 2026 (7:00 PM IST), with only 30 seats — registration closes on 9 November.
- Whole Genome Sequencing (WGS) Data Analysis Training
- Introduction to NGS and Sequencing QC
- Cancer Genomics with TCGA and GEO
👉 Explore all courses at www.biohelplearning.com
📩 Email: [email protected] | 📞 WhatsApp/Call: +91 78280 25951
