NEW YORK (GenomeWeb) – Researchers at Complete Genomics, a subsidiary of BGI, in collaboration with Reprogenetics and the NYU Facility Center, have used whole-genome sequencing to detect de novo single-base mutations in biopsies from early human embryos with high accuracy.
The method, which involves Complete Genomics' previously published long fragment read (LFR) technology for generating phased genomes, could have future applications in preimplantation genetic diagnosis (PGD) and circulating tumor cell analysis. The researchers published their results online in Genome Research today.
"We believe the approach could one day become a standard preimplantation diagnostic test for couples and individuals undergoing IVF," Santiago Munné, founder and director of Reprogenetics, told GenomeWeb. "The approach is extremely promising in its ability to detect de novo mutations with great sensitivity and specificity using very few embryo cells."
Reprogenetics, a PGD company based in Livingston, NJ, provided the embryos for the study and performed initial clinical testing on them.
"This is the first demonstration that de novo mutations can be detected in PGD with clinically relevant high specificity and sensitivity," said the study's senior author Rade Drmanac, CSO of Complete Genomics and senior vice president of BGI.
Current PGD testing, performed on one or several cells from an early-stage embryo, focuses on chromosomal aneuploidies, chromosomal rearrangements, and certain single-gene disorders but does not detect de novo single-base mutations. However, such mutations have been associated with a number of genetic disorders, including autism, intellectual disability, and schizophrenia.
For their study, the researchers took two biopsies of up to 10 cells from one five-day-old embryo and one biopsy from another embryo from the same couple. The couple had donated the embryos for research after undergoing IVF.
Complete Genomics generated three sequencing libraries from the three biopsies, using its LFR technology. For that, scientists split the DNA up into pools, where it is amplified by multiple displacement amplification and barcoded, so reads from each pool can be tracked back to the DNA fragment they originated from. They then sequenced the DNA on Complete Genomics' proprietary platform, followed by mapping and phasing.
One problem with sequencing highly amplified DNA is that the amplification process introduces large numbers of errors – on the order of 100,000 per sample – but the researchers were able to remove many of these spurious variants by putting them in the context of their haplotypes. After performing this haplotype analysis, 1,000 to 2,000 de novo single nucleotide variants remained in each embryo, about ten times more than the expected number of de novo mutations.
This number was further reduced by a filter that compares sequence data between several embryos from the same couple and by a well-count procedure. Drmanac explained that a real variant would be associated with a large number of different barcodes, whereas an error would only be covered by a few barcodes. The barcode counting process reduced the error rate by almost another 100-fold, he said, leaving just a few false positives per genome.
Overall, the scientists were able to detect and phase up to 82 percent of de novo single-base variants in the embryos, with a false-positive rate of about one error per gigabase, or fewer than 10 errors per embryo. In addition, they could detect small de novo deletions using the haplotype information.
While none of the de novo mutations affected any coding regions in one of the embryos, two mutations in coding regions of two genes were found in the other embryo. "However, it is unclear if there would be any detrimental effect to the health of a child born with these variants," the authors noted.
For introducing de novo mutation analysis as a clinical PGD test, much development still needs to go into the clinical interpretation of the results, Drmanac cautioned.
On the technical side, Complete Genomics is working on improving the sensitivity to be able to detect more than 82 percent of de novo mutations, and on lowering the bias introduced by the DNA amplification, according to Brock Peters, director of research at the company.
While the paper suggests the test and analysis currently takes several months to perform, Peters said that is no longer the case – sequencing for the publication was conducted several years ago, and the process has since improved.
Complete Genomics will continue to collaborate with Reprogenetics to optimize the technology and conduct further studies, Drmanac said. BGI will likely eventually commercialize the process as a test performed as part of its PGD services, following clinical studies.
In the meantime, Complete Genomics is exploring the approach for other applications, in particular circulating tumor cells. "It's more about the ability to collect these cells with high efficiency" than about sequencing them with high accuracy, Drmanac said. "We are working on this, and hopefully will get good results soon."