Widespread genotyping technique continuously misidentifies very uncommon pathogenic variants

SNP chips, a sort of DNA microarray continuously employed by client genomics corporations, are “extraordinarily unreliable” when figuring out clinically-relevant variants which can be uncommon among the many common inhabitants, in line with a retrospective examine not too long ago printed in The BMJ.

The investigation, carried out by researchers from the College of Exeter, reviewed sequencing information from almost 50,000 individuals of the U.Ok. Biobank and Private Genome Undertaking.

It discovered that SNP (single nucleotide polymorphism) chips had been largely correct when genotyping frequent variants discovered throughout the genome. When detecting a variant current amongst lower than .001% of the inhabitants, nevertheless, the tactic was extra seemingly to present a false constructive than a real constructive consequence.

These findings, a pair of the researchers wrote in an accompanying opinion article, could lead on sufferers to schedule invasive medical procedures, corresponding to bilateral prophylactic mastectomies, that they didn’t want.

“Though some client genomics corporations carry out sequencing to validate essential outcomes earlier than releasing them to shoppers, most shoppers additionally obtain their ‘uncooked’ SNP chip information for secondary evaluation, and this uncooked information nonetheless comprise these misguided outcomes,” Caroline Write, professor in genomic drugs on the College of Exeter, and Michael Weedon, affiliate professor in bioinformatics and human genetics on the College of Exeter, wrote within the opinion article.

“The implications of our findings are quite simple: SNP chips carry out poorly for detecting very uncommon genetic variants and the outcomes shouldn’t be utilized in scientific observe with out validation.”


When evaluating SNP chip genotyping to sequencing information for 49,908 U.Ok. Biobank individuals, the researchers discovered sensitivity, specificity, constructive predictive values and detrimental predictive values all to be above 99% for 108,574 frequent variants.

That efficiency dipped when the researchers checked out variants with a frequency under .001%, the place sensitivity dropped to 29.5% for one kind of array and 4.4% for one more. Optimistic predictive values fell among the many uncommon variants as effectively, the place 16.1% of 1 kind of microarray’s outcomes had been confirmed by sequencing and 9.4% of the opposite. These developments had been related among the many 21 individuals from the Private Genome Undertaking, for whom constructive predictive worth was 14%.

Whereas particularly reviewing variants throughout the BRCA1 and BRCA2 genes (chosen resulting from their relation to sure cancers), the SNP chips noticed 1,139 pathogenic or seemingly pathogenic variants, about 80% of which had been current amongst .01% of the U.Ok. Biobank inhabitants. The false positives had been once more frequent, driving sensitivity among the many U.Ok. Biobank individuals to 34.6% and constructive predictive worth to 4.2%. Equally, 20 of the 21 Private Genome Undertaking individuals had at the very least one false constructive of a uncommon pathogenic variant.


To conduct their retrospective evaluation, the researchers studied a subset of the U.Ok. Biobank’s full analysis cohort for whom each SNP chip genotyping and next-generation sequencing outcomes had been obtainable. These individuals had been recruited between 2006 and 2021, and 55% had been feminine. Equally, the crew reviewed datasets from the Private Genome Undertaking to incorporate a small pattern of people who had logged direct-to-consumer SNP chip and sequencing information.

The U.Ok. Biobank used SNP chip information genotyped utilizing the Utilized Biosystems UK Biobank Axiom Array (n = 45,871) and the Utilized Biosystems UK BiLEVE Axiom Array by Affymetrix (4,037), which the researchers famous have very related marker contents. Exome sequencing information for these had been generated for the mission by Regeneron.

For all 21 Private Genome Undertaking individuals, SNP chip information was offered by 23andMe utilizing Illumina arrays, whereas genome sequencing information was offered by Veritas Genetics.

With these information, the researchers in contrast the variants genotyped on the SNP chips to that particular person’s matching sequencing information.


Genomics corporations hit the patron well being house like a tidal wave, with gamers promising well being and wellness insights or new entry to genomic analysis information, whereas lawmakers and regulators scrambled to offer oversight. Clinicians are more and more studying to include information from these and different assessments into their workflows, whereas buyers have genomics corporations massive and small as funding – and even acquisition – targets.

Current strikes from main names recommend that the patron genomics well being market is not transferring full steam forward. For example, 23andMe laid off 100 staff within the wake of diminished demand and disappointing gross sales numbers – however has since raised new funding and introduced an SPAC merger. In the meantime Ancestry.com has backed away from its AncestryHealth providing lower than two years after its launch and lower 77 jobs within the course of.


“We recommend that, for variants which can be very uncommon within the inhabitants being examined, genotyping outcomes from SNP chips shouldn’t be routinely reported again to people or utilized in analysis with out validation. Clinicians and researchers ought to concentrate on the poor efficiency of SNP chips for genotyping very uncommon genetic variants to keep away from counseling sufferers inappropriately or investing restricted assets into investigating false associations with badly genotyped variants,” the researchers concluded.

You may also like...