Splicing and disease working group Purpose: Diagnosis of genetic disorders is hampered by large numbers of variants of uncertain significance (VUSs) identified through next-generation sequencing. Many such variants may disrupt normal RNA splicing. We examined effects on splicing of a large cohort of clinically identified variants and compared performance of bioinformatic splicing prediction tools commonly used in diagnostic laboratories. Methods: Two hundred fifty-seven variants (coding and noncoding) were referred for analysis across three laboratories. Blood RNA samples underwent targeted reverse transcription polymerase chain reaction (RT-PCR) analysis with Sanger sequencing of PCR products and agarose gel electrophoresis. Seventeen samples also underwent transcriptome-wide RNA sequencing with targeted splicing analysis based on Sashimi plot visualization. Bioinformatic splicing predictions were obtained using Alamut, HSF 3.1, and SpliceAI software. Results: Eighty-five variants (33%) were associated with abnormal splicing. The most frequent abnormality was upstream exon skipping (39/85 variants), which was most often associated with splice donor region variants. SpliceAI had greatest accuracy in predicting splicing abnormalities (0.91) and outperformed other tools in sensitivity and specificity. Conclusion: Splicing analysis of blood RNA identifies diagnostically important splicing abnormalities and clarifies functional effects of a significant proportion of VUSs. Bioinformatic predictions are improving but still make significant errors. RNA analysis should therefore be routinely considered in genetic disease diagnostics.

BackgroundPathogenicity predictors are integral to genomic variant interpretation but, despite their widespread usage, an independent validation of performance using a clinically relevant dataset has not been undertaken.MethodsWe derive two validation datasets: an ‘open’ dataset containing variants extracted from publicly available databases, similar to those commonly applied in previous benchmarking exercises, and a ‘clinically representative’ dataset containing variants identified through research/diagnostic exome and panel sequencing. Using these datasets, we evaluate the performance of three recent meta-predictors, REVEL, GAVIN and ClinPred, and compare their performance against two commonly used in silico tools, SIFT and PolyPhen-2.ResultsAlthough the newer meta-predictors outperform the older tools, the performance of all pathogenicity predictors is substantially lower in the clinically representative dataset. Using our clinically relevant dataset, REVEL performed best with an area under the receiver operating characteristic curve of 0.82. Using a concordance-based approach based on a consensus of multiple tools reduces the performance due to both discordance between tools and false concordance where tools make common misclassification. Analysis of tool feature usage may give an insight into the tool performance and misclassification.ConclusionOur results support the adoption of meta-predictors over traditional in silico tools, but do not support a consensus-based approach as in current practice.

ObjectiveRare genetic disorders resulting in prenatal or neonatal death are genetically heterogeneous, but testing is often limited by the availability of fetal DNA, leaving couples without a potential prenatal test for future pregnancies. We describe our novel strategy of exome sequencing parental DNA samples to diagnose recessive monogenic disorders in an audit of the first 50 couples referred.MethodExome sequencing was carried out in a consecutive series of 50 couples who had 1 or more pregnancies affected with a lethal or prenatal‐onset disorder. In all cases, there was insufficient DNA for exome sequencing of the affected fetus. Heterozygous rare variants (MAF < 0.001) in the same gene in both parents were selected for analysis. Likely, disease‐causing variants were tested in fetal DNA to confirm co‐segregation.ResultsParental exome analysis identified heterozygous pathogenic (or likely pathogenic) variants in 24 different genes in 26/50 couples (52%). Where 2 or more fetuses were affected, a genetic diagnosis was obtained in 18/29 cases (62%). In most cases, the clinical features were typical of the disorder, but in others, they result from a hypomorphic variant or represent the most severe form of a variable phenotypic spectrum.ConclusionWe conclude that exome sequencing of parental samples is a powerful strategy with high clinical utility for the genetic diagnosis of lethal or prenatal‐onset recessive disorders. © 2017 The Authors Prenatal Diagnosis published by John Wiley & Sons Ltd.

Recent studies have identified both recessive and dominant forms of mitochondrial disease that result from ATAD3A variants. The recessive form includes subjects with biallelic deletions mediated by non-allelic homologous recombination. We report five unrelated neonates with a lethal metabolic disorder characterized by cardiomyopathy, corneal opacities, encephalopathy, hypotonia, and seizures in whom a monoallelic reciprocal duplication at the ATAD3 locus was identified. Analysis of the breakpoint junction fragment indicated that these 67 kb heterozygous duplications were likely mediated by non-allelic homologous recombination at regions of high sequence identity in ATAD3A exon 11 and ATAD3C exon 7. At the recombinant junction, the duplication allele produces a fusion gene derived from ATAD3A and ATAD3C, the protein product of which lacks key functional residues. Analysis of fibroblasts derived from two affected individuals shows that the fusion gene product is expressed and stable. These cells display perturbed cholesterol and mitochondrial DNA organization similar to that observed for individuals with severe ATAD3A deficiency. We hypothesize that the fusion protein acts through a dominant-negative mechanism to cause this fatal mitochondrial disorder. Our data delineate a molecular diagnosis for this disorder, extend the clinical spectrum associated with structural variation at the ATAD3 locus, and identify a third mutational mechanism for ATAD3 gene cluster variants. These results further affirm structural variant mutagenesis mechanisms in sporadic disease traits, emphasize the importance of copy number analysis in molecular genomic diagnosis, and highlight some of the challenges of detecting and interpreting clinically relevant rare gene rearrangements from next-generation sequencing data.

ContextTo date, penetrance figures for medullary thyroid cancer (MTC) for variants in rearranged during transfection (RET) have been estimated from families ascertained because of the presence of MTC.ObjectiveTo gain estimates of penetrance, unbiased by ascertainment, we analyzed 61 RET mutations assigned as disease causing by the American Thyroid Association (ATA) in population whole-exome sequencing data.DesignFor the 61 RET mutations, we used analyses of the observed allele frequencies in ∼51,000 individuals from the Exome Aggregation Consortium (ExAC) database that were not contributed via The Cancer Genome Atlas (TCGA; non-TCGA ExAC), assuming lifetime penetrance for MTC of 90%, 50%, and unbounded.SettingPopulation-based.ResultsTen of 61 ATA disease-causing RET mutations were present in the non-TCGA ExAC population with observed frequency consistent with penetrance for MTC of >90%. For p.Val804Met, the lifetime penetrance for MTC, estimated from the allele frequency observed, was 4% [95% confidence interval (CI), 0.9% to 8%].ConclusionsBased on penetrance analysis in carrier relatives of p.Val804Met-positive cases of MTC, p.Val804Met is currently understood to have high-lifetime penetrance for MTC (87% by age 70), albeit of later onset of MTC than other RET mutations. Given our unbiased estimate of penetrance for RET p.Val804Met of 4% (95% CI, 0.9% to 8%), the current recommendation by the ATA of prophylactic thyroidectomy as standard for all RET mutation carriers is likely inappropriate.

Background ATPase family AAA-domain containing protein 3A (ATAD3A) is a nuclear-encoded mitochondrial membrane-anchored protein involved in diverse processes including mitochondrial dynamics, mitochondrial DNA organization, and cholesterol metabolism. Biallelic deletions (null), recessive missense variants (hypomorph), and heterozygous missense variants or duplications (antimorph) in ATAD3A lead to neurological syndromes in humans. Methods To expand the mutational spectrum of ATAD3A variants and to provide functional interpretation of missense alleles in trans to deletion alleles, we performed exome sequencing for identification of single nucleotide variants (SNVs) and copy number variants (CNVs) in ATAD3A in individuals with neurological and mitochondrial phenotypes. A Drosophila Atad3a Gal4 knockin-null allele was generated using CRISPR-Cas9 genome editing technology to aid the interpretation of variants. Results We report 13 individuals from 8 unrelated families with biallelic ATAD3A variants. The variants included four missense variants inherited in trans to loss-of-function alleles (p.(Leu77Val), p.(Phe50Leu), p.(Arg170Trp), p.(Gly236Val)), a homozygous missense variant p.(Arg327Pro), and a heterozygous non-frameshift indel p.(Lys568del). Affected individuals exhibited findings previously associated with ATAD3A pathogenic variation, including developmental delay, hypotonia, congenital cataracts, hypertrophic cardiomyopathy, and cerebellar atrophy. Drosophila studies indicated that Phe50Leu, Gly236Val, Arg327Pro, and Lys568del are severe loss-of-function alleles leading to early developmental lethality. Further, we showed that Phe50Leu, Gly236Val, and Arg327Pro cause neurogenesis defects. On the contrary, Leu77Val and Arg170Trp are partial loss-of-function alleles that cause progressive locomotion defects and whose expression leads to an increase in autophagy and mitophagy in adult muscles. Conclusion Our findings expand the allelic spectrum of ATAD3A variants and exemplify the use of a functional assay in Drosophila to aid variant interpretation.

Proteins anchored to the cell surface via glycosylphosphatidylinositol (GPI) play various key roles in the human body, particularly in development and neurogenesis. As such, many developmental disorders are caused by mutations in genes involved in the GPI biosynthesis and remodeling pathway. We describe ten unrelated families with bi-allelic mutations in PIGB, a gene that encodes phosphatidylinositol glycan class B, which transfers the third mannose to the GPI. Ten different PIGB variants were found in these individuals. Flow cytometric analysis of blood cells and fibroblasts from the affected individuals showed decreased cell surface presence of GPI-anchored proteins. Most of the affected individuals have global developmental and/or intellectual delay, all had seizures, two had polymicrogyria, and four had a peripheral neuropathy. Eight children passed away before four years old. Two of them had a clinical diagnosis of DOORS syndrome (deafness, onychodystrophy, osteodystrophy, mental retardation, and seizures), a condition that includes sensorineural deafness, shortened terminal phalanges with small finger and toenails, intellectual disability, and seizures; this condition overlaps with the severe phenotypes associated with inherited GPI deficiency. Most individuals tested showed elevated alkaline phosphatase, which is a characteristic of the inherited GPI deficiency but not DOORS syndrome. It is notable that two severely affected individuals showed 2-oxoglutaric aciduria, which can be seen in DOORS syndrome, suggesting that severe cases of inherited GPI deficiency and DOORS syndrome might share some molecular pathway disruptions.

IntroductionSpermatogenesis is a highly complex process involving several thousand genes, only a minority of which have been studied in infertile men. In a previous study, we identified a number of Copy Number Variants (CNVs) by high-resolution array-Comparative Genomic Hybridization (a-CGH) analysis of the X chromosome, including 16 patient-specific X chromosome-linked gains. Of these, five gains (DUP1A, DUP5, DUP20, DUP26 and DUP40) were selected for further analysis to evaluate their clinical significance.Materials and MethodsThe copy number state of the five selected loci was analyzed by quantitative-PCR on a total of 276 idiopathic infertile patients and 327 controls in a conventional case-control setting (199 subjects belonged to the previous a-CGH study). For one interesting locus (intersecting DUP1A) additional 338 subjects were analyzed.Results and DiscussionAll gains were confirmed as patient-specific and the difference in duplication load between patients and controls is significant (p = 1.65×10−4). Two of the CNVs are private variants, whereas 3 are found recurrently in patients and none of the controls. These CNVs include, or are in close proximity to, genes with testis-specific expression. DUP1A, mapping to the PAR1, is found at the highest frequency (1.4%) that was significantly different from controls (0%) (p = 0.047 after Bonferroni correction). Two mechanisms are proposed by which DUP1A may cause spermatogenic failure: i) by affecting the correct regulation of a gene with potential role in spermatogenesis; ii) by disturbing recombination between PAR1 regions during meiosis. This study allowed the identification of novel spermatogenesis candidate genes linked to the 5 CNVs and the discovery of the first recurrent, X-linked gain with potential clinical relevance.