Biology Atlas

Map / Assays

assay · intermediate · partly written

Whole-genome sequencing

Whole-genome sequencing reads essentially the whole of a person's genome as short reads, which are mapped to a reference and turned into variant calls.

Also called: WGS

Error rate and what it means more

Illumina states error rates below 1 in 1,000 per base (Q30); substitutions dominate and errors depend on sequence context, so they are not random noise to average away. No genome-wide error rate or caller F1 score from an opened primary source was found, so none is given. In a haploid test sample, raw genotype error was about 1 per 10 to 15 kb, falling to about 1 per 100 to 200 kb after filters. 12

Reported, not yet replicated

Genome, exome and array compared more

The genome sees noncoding regions where most GWAS signal lies; the exome targets coding regions only; an array measures a chosen set of tag SNPs. The chain found no head-to-head cost or sensitivity comparison of the three in an opened source. The cost in this record is the one already filed and was not re-derived. 34

Reported, not yet replicated

Question answered

Which positions in the whole genome differ from the reference in this person, coding and noncoding, including variants a capture-based exome never targets. 5

Reported, not yet replicated

Readout physical

Millions of short reads of about 100 to 150 letters (a count of reads and their base qualities), placed on the reference and then called as variants. 1

Reported, not yet replicated

Not measured

Repeats and near-identical gene copies where reads cannot be placed uniquely: a missing variant there can mean could not see. The newest benchmark covers 92.2% of autosomal bases, and a short-read call set missed 16,615 SNVs against it versus 1,960 against the older one. 5

Reported, not yet replicated

Setting

Where in cascade

Cost

Per sample

221 USD per 30x human genome (external core rate, batch of 24) (University of Minnesota Genomics Center page, undated: PCR-free preparation, 30x sequencing, QC and variant calling. Core rate; excludes downstream analysis and storage. Other sources quote other scopes (see data_availability).; as of 2026-11-17)6

ML role

Does

Neural-network callers such as DeepVariant classify candidate sites from a six-channel pileup image; statistical callers use haplotype assembly and a pair-HMM. 78

Reported, not yet replicated

Cannot

No caller fixes a read that was placed on the wrong gene copy; accuracy is a property of the region as much as the caller. 5

Our inference

Data availability

University of Minnesota Genomics Center purchase

A 30x human genome including PCR-free library preparation, sequencing, QC and variant calling. 6

Caveats. Core rate for a standard product, undated page; University of Minnesota users pay less.

Price. 221 USD per 30x human genome (external rate, batch of 24) (UMN Genomics Center page, undated; core rate.; as of 2026-11-17)6

See: University of Minnesota Genomics Center

Reported, not yet replicated

Fully loaded in-house sequencing (worked example) in-house

A trade article quotes a rate-card figure of about $374 per genome and a fully loaded in-house model that adds capital, storage, analyst time and repeat runs. 910

Caveats. Secondary article with worked examples for a 1,000-genome-a-year programme; its own FAQ conflicts with its numbers. The national fact sheet excludes QA/QC and analysis from its cost series.

Price. 1,168 to 1,939 USD per genome, fully loaded, 1,000 genomes a year (Lab Manager worked example (secondary); capital, storage, analyst time and repeats included.; as of 2026)9

Reported, not yet replicated

Still open here

Not yet written: Sample prep, Conditions, Throughput class, Throughput, Time per run, Engineerable variants, Use cases, Before after, Capital, Emerging.

Linked from

Sources

  1. Tan G, Opitz L, Schlapbach R, et al.. Long fragments achieve lower base quality in Illumina paired-end sequencing. Sci Rep (2019). review; opened 2026-11-17 all records citing this
  2. li-2014-variant-calling-artifacts.. preprint, not peer reviewed; opened 2026-11-17 all records citing this
  3. Backman JD, Li AH, Marcketta A, et al.. Exome sequencing and analysis of 454,787 UK Biobank participants. Nature (2021). paper; opened 2026-11-17 all records citing this
  4. Klein, R., Zeiss, C., Chew, E., et al.. Complement Factor H Polymorphism in Age-Related Macular Degeneration. Science (2005). paper; opened 2026-11-17 all records citing this
  5. Wagner J, Olson ND, Harris L, et al.. Benchmarking challenging small variants with linked and long reads. Cell Genom (2022). paper; opened 2026-11-17 all records citing this
  6. University of Minnesota Genomics Center: human whole-genome sequencing. University of Minnesota (2026). organisation-page, not peer reviewed; opened 2026-11-17 all records citing this
  7. Google DeepVariant team. DeepVariant details (release r1.6.1). GitHub documentation. web, not peer reviewed; opened 2026-11-17 all records citing this
  8. Illumina. DRAGEN small variant calling algorithm. Illumina support documentation. vendor-page, not peer reviewed; opened 2026-11-17 all records citing this
  9. How much does NGS cost? Budgeting instruments, reagents and sequencing (Lab Manager). Lab Manager (2026). news, not peer reviewed; opened 2026-11-17 all records citing this
  10. NHGRI: the cost of sequencing a human genome (fact sheet). National Human Genome Research Institute (2026). organisation-page, not peer reviewed; opened 2026-11-17 all records citing this
  11. Amit's dictation of 2026-11-17: what the atlas must contain. Kept in the atlas repository; not published.
Where it sits on the map
History of this record (3)

This record has been changed 3 times, on the dates below. The note kept against each change is written for the people who maintain the atlas, in their own shorthand, so it is not shown here; it is in the record itself.

Status
partly written
Last updated
2026-11-17
Record
content/assays/whole-genome-sequencing.json