WGS, WES and genotyping: how to choose a genetic testing method
Human DNA can be analyzed by several technologically different methods, and the choice of method depends on the goal of the analysis and the type of variants that need to be detected. These include whole-genome sequencing, exome sequencing, genotyping, PCR and others. Genetic tests that assess predisposition to the most common conditions most often use genotyping, since it is the most optimal in terms of informativeness and cost. Using PCR or whole-exome analysis is not appropriate here, because analyzing a single gene does not reflect the full picture of predisposition to a range of conditions. That is why genotyping and whole-genome analysis are used to calculate a polygenic predisposition score.
Genome, exome and a fixed set of variants: what each method analyzes
The choice of analysis method depends on which part of the DNA needs to be examined and which variants need to be detected in it. Some methods determine the DNA sequence across the whole genome or in particular regions of it, while others check the genotype at predefined positions. This determines which variants a method is able to detect. Together with the accuracy and reproducibility of the analysis, this defines the tasks a method can be used for.
Whole genome sequencing (WGS)
WGS makes it possible to examine the sequence of almost the entire genome — about 3.2 billion base pairs, the structural units that make up DNA. Sequencing covers all genetic material: the coding regions of genes, introns (non-coding regions within genes) and regulatory sequences that influence gene activity. This method is not limited to a predefined set of positions (Kockum, Huang, & Stridh, Current Protocols, 2023).
Whole exome sequencing (WES)
WES determines the sequence of the exome — the set of all exons, i.e. the protein-coding regions of genes, which together make up roughly 1–2% of the genome. The DNA fragments corresponding to exons are selectively captured before sequencing; the rest of the genome is not included in the analysis (Kockum, Huang, & Stridh, Current Protocols, 2023).
Genotyping (SNP array)
The method determines the genotype at a predefined set of known positions using hybridization probes designed for each of them. The DNA sequence itself is not determined. The widely used Illumina Global Screening Array platform, for example, covers about 700,000 markers (Illumina, Infinium Global Screening Array-24 v3.0 Data Sheet, 2020).
Which classes of variants each method can detect
WGS: coverage without predefined limits
WGS is not limited to a predefined set of positions, so it gives access to variants in any part of the genome: rare variants outside the exome, variants in non-coding regulatory regions, and structural variants. The ability to detect structural variants depends on the sequencing technology and the data analysis algorithm. An analysis of whole-genome data from 150,119 UK Biobank participants showed that direct sequencing makes it possible to establish associations for rare variants whose imputation accuracy is limited by their low representation in reference panels (Halldorsson et al., Nature, 2022).
WES: read depth of coding regions
WES focuses sequencing on the exome, so it provides greater read depth of coding regions with a smaller volume of data than whole-genome sequencing at comparable depth (Kockum, Huang, & Stridh, Current Protocols, 2023). In the diagnosis of hereditary conditions this is reflected in the balance between outcome and cost: according to a comparative analysis of Mendelian disorders, WGS provides a higher diagnostic yield, while WES with reanalysis of the data gives a lower cost of investigation, and the choice between the methods is determined by the clinical scenario and the available resources (Ewans et al., Eur J Hum Genet, 2022).
Genotyping: large-scale coverage of known positions
PRS aggregates the contribution of a large number of genetic variants, mostly common in the population, each of which usually has a small individual effect. Such a calculation requires the genotype to be determined reliably and reproducibly at a large number of already known positions. A microarray genotypes a fixed set of positions, while genotypes at neighboring positions that were not analyzed directly are reconstructed by imputation (statistical completion of data based on reference panels) (Kockum, Huang, & Stridh, Current Protocols, 2023).
More about how genetic data is structured — in the article Types of genetic data: what lies behind SNPs, haplotypes and polygenic profiles.
Coverage, read depth and genotyping accuracy
Sequencing and genotyping produce different types of data, so the reliability of the result is assessed using different parameters. For WGS and WES the main parameters are coverage (what fraction of the target region was included in the analysis) and read depth (how many times each position was independently determined). Depth affects the reliability of calling a variant, especially at heterozygous positions (where the variants inherited from the parents are different).
For a microarray, other quality metrics are used, in particular call rate and reproducibility. Call rate is the fraction of array markers that were successfully read in a given sample. It describes the completeness of a single measurement. Reproducibility shows how consistent the genotyping results for the same sample are between repeated runs, i.e. it describes the stability of the platform. According to the manufacturer, for the Illumina Global Screening Array v3.0 the average call rate is 99.5% against a specification of over 99.0%, and reproducibility is 99.99% against a specification of over 99.90% (Illumina, Infinium Global Screening Array-24 v3.0 Data Sheet, 2020).
The quality of the next step (imputation) is assessed with separate metrics, in particular r², which indicate the expected reliability of an imputed genotype. The specific value depends on the imputation algorithm, the size and composition of the reference panel and the way the data is processed.
More about the formats in which genetic data is stored — in the article Genetic data formats: FASTQ, PLINK, VCF and others.
Read depth is an adjustable parameter, and the balance between direct genotype determination and imputation depends on it. Low-pass WGS is performed at lower depth than standard whole-genome sequencing, and genotypes at insufficiently covered positions are reconstructed by imputation. For calculating PRS, this approach can give results comparable to microarray genotyping (Li, Mazur, Berisa, & Pickrell, Genome Research, 2021).
Applications of WGS, WES and microarray
The scope of a method's application is determined both by its technical capabilities and by the circumstances of the specific study or clinical case:
-
WGS is used in genome research, to search for rare and structural variants, and also in the clinical diagnosis of hereditary conditions when the causal variant may lie outside the protein-coding regions.
-
WES is used in the diagnosis of monogenic hereditary diseases, where the search is focused on coding sequences.
-
Genotyping with imputation is used for large-scale polygenic analysis, since it makes it possible to reliably analyze a large number of known markers and to supplement the result with reference-panel data.
The choice between WES and WGS in clinical practice is determined by the study scenario and the available resources (Ewans et al., Eur J Hum Genet, 2022).
Microarray genotyping in Apixmed Prism
Genotyping makes it possible to analyze a large number of positions using a standardized protocol and with high reproducibility of the result. A predisposition estimate is built on a combination of many markers, so the reliability of calling each of them affects the final result.
A saliva DNA sample is processed by a laboratory on a platform that genotypes about 700,000 genetic markers. The indicators in the report are calculated from the analyzed variants and, where necessary, imputation data, using data obtained from the relevant GWAS studies. This is how the results of Apixmed Prism genetic tests and panels are generated.
The scale of coverage and the depth of interpretation work together here: the more markers are included in the analysis and the more complete the evidence base for each of them, the more detailed the picture of genetic predisposition the report provides.
Learn more about Apixmed Prism genetic testing →
WGS, WES and microarray: answers to common questions
Can whole-genome sequencing be done later if a genotyping-based test was done first?
Yes. These are separate methods, and genotyping results do not limit the ability to undergo sequencing later. It requires a new sample, because the methods use different DNA preparation protocols.
Does choosing a genotyping chip mean that part of the genetic information remains inaccessible?
A genotyping chip analyzes a predefined set of positions. Questions that go beyond this set — for example, searching for rare variants not covered by the chip's design — are addressed using WES or WGS.
Does WGS automatically mean a more accurate genetic test?
No. WGS provides broader coverage of the DNA sequence, but the accuracy and usefulness of the result depend on the specific class of variants, the sequencing depth, the analysis algorithms and the task at hand.
Can the results of whole-genome sequencing and a genotyping-based test be compared directly?
Directly — no, because the methods determine the genotype differently and cover different regions of the genome. Specific indicators can be compared if they are calculated using the same PRS models and comparable reference panels.
The results of a genetic test are not a diagnosis and not a substitute for a consultation with a doctor. The Apixmed Prism report provides genetic context that complements the results of examinations and helps make decisions together with a doctor.
Sources
1. Kockum, I., Huang, J., & Stridh, P. (2023). Overview of genotyping technologies and methods. Current
Protocols, 3(4), e727. https://doi.org/10.1002/cpz1.727
2. Halldorsson, B. V., et al. (2022). The sequences of 150,119 genomes in the UK Biobank. Nature, 607(7920), 732–740. https://doi.org/10.1038/s41586-022-04965-x
3. Li, J. H., Mazur, C. A., Berisa, T., & Pickrell, J. K. (2021). Low-pass sequencing increases the power of GWAS and decreases measurement error of polygenic risk scores compared to genotyping arrays. Genome Research, 31(4), 529–537. https://doi.org/10.1101/gr.267229.120
4. Illumina, Inc. (2020). Infinium Global Screening Array-24 v3.0 BeadChip: Data Sheet (document 370-2016-016-G).
5. Ewans, L. J., Minoche, A. E., Schofield, D., et al. (2022). Whole exome and genome sequencing in mendelian disorders: a diagnostic and health economic analysis. European Journal of Human Genetics, 30(10), 1121–1131. https://doi.org/10.1038/s41431-022-01162-2
