Sequencing a genome produces a result for one person and a dataset for everyone who follows. Building the second from the first requires consent, storage and a governance arrangement that survives the study that created it. Genomics England has run that experiment once, and has now extended it to newborn screening.

From a project to a library

The 100,000 Genomes Project sequenced participants with rare disease and cancer, and the diagnostic findings from its pilot were published in the New England Journal of Medicine in November 2021. That pilot covered 4,660 participants from 2,183 families across 161 disorders.

The reported diagnostic yield was 25 per cent of probands, rising to 35 per cent where a monogenic cause was likely and falling to 11 per cent where a complex cause was likely. Fourteen per cent of diagnoses depended on combining research and automated analysis, which mattered for variants that other methods would miss.

The programme delivered more than diagnoses. Cohort-wide burden testing across 57,000 genomes identified three new disease genes and 19 new associations, and a quarter of the genetic diagnoses made had immediate implications for clinical decisions for the patients or their relatives.

The dataset outlives the project that produced it. Genomics England maintains the National Genomic Research Library, drawing on participants recruited through the project and through NHS Genomic Medicine Centres, and makes it available to approved researchers.

The library's terms are narrower than the sequencing it holds. Only data that participants have consented to share are available to approved researchers, and participants can check what they have consented to and change their minds. Results from the project are still being returned years after recruitment closed.

That matters for what a national dataset is: not a fixed file but a set of permissions granted at a moment, maintained and re-asked over time. The project sequenced genomes from around 85,000 NHS patients affected by rare disease or cancer, reaching 100,000 genomes by December 2018.

The Generation Study

The Generation Study is a research programme run by Genomics England with the NHS that offers whole genome sequencing to newborn babies. It tests for changes in genes linked to more than 200 rare genetic conditions, concentrating on conditions not currently included in the UK newborn screening programme and where earlier identification could allow earlier support, monitoring or treatment.

Parents sign up during pregnancy, through participating NHS trusts in England, and a blood sample is collected shortly after birth from the umbilical cord or by heel prick. Sequencing looks only for specific variants known to cause one of the conditions on the study list. Variants of uncertain significance are not screened for, which the programme states is intended to limit false positives.

It is a screening test rather than a diagnostic one. It examines a defined set of genes rather than searching broadly for a cause, and for most conditions a follow-up non-genomic test can confirm whether the baby is affected. In the UK all newborns are already offered a blood spot test around day five of life, which measures biochemical markers for nine conditions. That remains the standard of care, and the study runs alongside it rather than replacing it.

The study's stated target is 100,000 newborn participants. The conditions it screens for were chosen on the basis that a result could lead to action: the point is to find cases where knowing sooner changes what happens next.

Condition selection follows four principles developed for the study, and every condition screened has been approved by expert clinicians and NHS England. The Generation Study has already produced a case that illustrates how a result is meant to work: a newborn was found to have a rare growth condition through the sequencing, and the finding reached the family through the study pathway.

Because the study concerns babies, consent is given by parents on the child's behalf and the data are retained long term. That is the part of the design carrying the most weight and the least evidence, because no comparable programme has run long enough to show how families regard the decision decades later. Consent is treated as ongoing rather than as a single event: participants can withdraw at any time, and when a child reaches around sixteen they are asked whether they wish to remain in the study. That provision tests whether the original decision still holds once the person it concerns can make it themselves.

What the evidence shows so far

Recruitment is approaching completion. As of July 2026, more than 75,000 parents had been recruited and more than 56,000 results returned, with recruitment expected to finish in early 2027 at the planned target of 100,000 participants and final results due back to families in spring 2027.

More than 180 participants had received a result where a condition was suspected and were referred for specialist NHS care. That number is the study's principal clinical output so far, and it is small relative to the cohort, which is what a screening test for rare conditions produces by design.

The proportions explain that. Around 99 per cent of participants receive a result of no condition suspected, and around 1 per cent receive a result where a condition is suspected. For that 1 per cent, Genomics England returns the result to the NHS through regional result coordinators, who identify the appropriate specialist team; that team contacts the family to arrange confirmatory testing and begin care.

The return pathway is slower than the screening model implies. In the early stages results take a few months to reach families, with the study expecting that timeline to fall to two to three weeks. A screening test that identifies a treatable condition in a newborn is only as useful as the interval between the sample and the family knowing the answer.

The study is being evaluated, not merely operated. Its purpose is to establish whether sequencing newborns at scale is feasible and whether it improves outcomes, and those questions are answered by analysis after recruitment closes rather than by the volume of participants enrolled.

What is not yet known

The Generation Study will end without answering the question that governs whether its approach is adopted. A research programme can show that conditions are detected and that families are referred. It cannot by itself show that earlier detection changed the course of a condition, because the study was not designed with an untreated comparison group, and because the outcomes that matter may take years to appear.

What is not yet known is whether the evidence collected will be judged sufficient for newborn screening policy. Those decisions belong to the UK National Screening Committee and NHS England, and the study's role is to inform them rather than to make them. Whether the screening panel changes as a result is not yet known.

Summary

Genomics England has demonstrated that national-scale sequencing can produce diagnoses and a durable research resource. Whether the same model should be applied to every newborn is the question now under test.

  • The library came first. The 100,000 Genomes Project pilot reported a 25 per cent diagnostic yield across 4,660 participants.
  • The study targets screening. The Generation Study offers whole genome sequencing to newborns for more than 200 rare conditions.
  • Recruitment is nearly complete. More than 75,000 parents recruited and more than 56,000 results returned as of July 2026, against a target of 100,000.
  • Referrals are the clinical output. More than 180 participants with a suspected condition have been referred for specialist NHS care.
  • The policy question is separate. Decisions on newborn screening rest with the UK National Screening Committee and NHS England, not with the study.

Whether sequencing newborns improves outcomes enough to justify a national programme is not yet known.

Research direction and editorial oversight: Juno Uni Hera. AI-assisted research and drafting; human review is ongoing.

Sources