Validation of a next generation sequencing-based preimplantation genetic screening assay for the calling of triploidy and uniparental isodisomy

Abstract

We have previously validated the FAST-SeqS NGS-based PGS technology for the calling of whole chromosome and segmental aneuploidy (1,2). This technology captures sequence elements that cover ten to twenty thousand polymorphic sites. The relative representation of alleles at such sites can be used to call additional forms of chromosomal abnormality associated with negative reproductive outcomes, including triploidy, the presence of an extra set of chromosomes, and uniparental isodisomy (UPiD), the presence of two copies of the same chromosome from a single parent. To validate FAST-SeqS for the calling of triploidy and uniparental isodisomy. Samples from five triploid (69,XXY and 69,XXX), three UPiD (UPiD(7), UPiD(8), UPiD(1-22,X)) and eight diploid (47,XY,+9, 47,XY,+13, 47,XX,+18, 47,XY,+21, 47,XXY, 47,XYY, 46,XY, 46,XX) cell lines were run through FAST-SeqS. The resulting ploidy and UPiD calls were compared to the expected karyotypes to estimate analytical sensitivity and specificity. Next, to evaluate the dependence of triploidy and UPiD calling on cell count, cells from the lines noted above were run through FAST-SeqS at one, two, five or ten cells. Finally, to estimate the rate of triploidy and UPiD in blastocysts, we processed the FAST-SeqS data for 7365 embryos utilizing our validated analysis pipeline. All 76 triploid and 36 UPiD samples were called correctly, leading to triploidy and UPiD sensitivity estimates (95% confidence intervals) of 100% (95.8-100%) and 100% (98.4-100%), respectively. There were no false positive triploidy or UPiD calls for 210 known diploid samples, yielding triploidy and UPiD specificity estimates of 100% (98.2-100%) and 100% (98.2-100%), respectively. Across the two, five, and ten cell known triploid/UPiD samples (n=43, 53, 105), we observed zero false negative triploidy or UPiD calls. Similarly, no false positive triploidy or UPiD calls were made for the one, two, and five cell diploid samples (n=64, 132, 210). Amongst 7365 blastocyst samples, 69 were called triploid and 6 UPiD, yielding rate estimates of 0.94% (0.71-1.2%) and 0.08% (0.04-0.18%), respectively, values that are consistent with previous estimates (3-4). Of note, 33 of the triploid embryos were 69,XXX and therefore would not have been identifiable by an atypical sex chromosome copy number ratio. We have validated an enhanced version of the FAST-SeqS PGS assay for the accurate calling of triploidy and uniparental isodisomy, forms of chromosomal abnormality that are often not detected on other PGS platforms.