GeneSplicer: a new computational method for splice site prediction

Ocular manifestations of CHARGE syndrome in a pediatric cohort with genotype/phenotype analysis

CHARGE syndrome is a rare multi-system condition associated with CHD7 variants. However, ocular manifestations and particularly ophthalmic genotype-phenotype associations, are not well-studied. This study evaluated ocular manifestations and genotype-phenotype associations in pediatric patients with CHARGE syndrome. A retrospective chart review included pediatric patients under 20 years-old with clinical diagnosis of CHARGE syndrome and documented ophthalmic examination. Demographics, genetic testing, and ocular findings were collected. Comprehensive literature review enhanced the genotype-phenotype analysis. Forty-two patients (20 male) underwent eye examination at an average age of 9.45 ± 6.52 years-old. Thirty-nine (93%) had ophthalmic manifestations in at least one eye. Optic nerve/chorioretinal colobomas were most common (38 patients), followed by microphthalmia (13), cataract (6), and iris colobomas (4). Extraocular findings included strabismus (32 patients), nasolacrimal duct obstructions (11, 5 with punctal agenesis), and cranial nerve VII palsy (10). Genotype-phenotype analyses (27 patients) showed variability in ocular phenotypes without association to location or variant types. Splicing (10 patients) and frameshift (10) variants were most prevalent. Patients with CHARGE syndrome may present with a myriad of ophthalmic manifestations. There is limited data regarding genotype-phenotype correlations and additional studies are needed.

Genome sequencing identifies biallelic variants in SCLT1 in a patient with syndromic nephronophthisis: Reflections on the SCLT1-related ciliopathy spectrum

Ciliopathies represent a major category of rare multisystem disease. Arriving at a specific diagnosis for a given patient is challenged by the significant genetic and clinical heterogeneity of these conditions. We report the outcome of the diagnostic odyssey of a child with obesity, renal, and retinal disease. Genome sequencing identified biallelic splice site variants in sodium channel and clathrin linker 1 (SCLT1), an emerging ciliopathy gene. We review the literature on all patients reported with biallelic SCLT1 variants highlighting a frequent clinical presentation that overlaps Bardet-Biedl and Senior-Loken syndromes. We also discuss current concepts in syndrome designation in light of these data.

Clinical and Genetic Findings in a Cohort of Patients with PRPF31-associated Retinal Dystrophy

PURPOSE

The purpose of our study was to assess the phenotypic and genotypic spectrum in a large cohort of patients with PRPF31-associated retinal dystrophy.

DESIGN

Retrospective cohort study METHODS: In this retrospective chart review study, we collected cross-sectional data on the phenotype and genotype of patients with PRPF31-associated retinal dystrophy from the clinics for inherited retinal dystrophies at the University of Tuebingen and the local RetDis database and biobank. Patients underwent thorough ophthalmological examinations and genetic testing.

RESULTS

Eighty-six patients from 61 families were available for clinical assessment, while genomic DNA was available for 111 individuals (index patients and family members). Fifty-three different disease-associated variants were observed in our cohort. Point mutations were the most common class. All but two patients exhibited features of a typical Retinitis pigmentosa (RP). One patient showed a cone-rod-dystrophy pattern. One mutation carrier revealed no signs of a retinal dystrophy. There was a statistically significant better visual acuity for patients with large deletions in the 20-39 age group. Cystoid macular edema was common in those with preserved central retina and showed an association with female sex.

CONCLUSION

Our study confirms high phenotypic variability in disease onset and age at which legal blindness is reached in PRPF31-linked RP. Non-penetrance is commonly documented in family history, although poorly represented in our study, possibly indicating that true asymptomatic mutation carriers are rare if followed-up over lifetime with thorough ophthalmologic workup.

The Influence of a Genetic Variant in CCDC78 on LMNA-Associated Skeletal Muscle Disease

Mutations in the LMNA gene-encoding A-type lamins can cause Limb-Girdle muscular dystrophy Type 1B (LGMD1B). This disease presents with weakness and wasting of the proximal skeletal muscles and has a variable age of onset and disease severity. This variability has been attributed to genetic background differences among individuals; however, such variants have not been well characterized. To identify such variants, we investigated a multigeneration family in which affected individuals are diagnosed with LGMD1B. The primary genetic cause of LGMD1B in this family is a dominant mutation that activates a cryptic splice site, leading to a five-nucleotide deletion in the mature mRNA. This results in a frame shift and a premature stop in translation. Skeletal muscle biopsies from the family members showed dystrophic features of variable severity, with the muscle fibers of some family members possessing cores, regions of sarcomeric disruption, and a paucity of mitochondria, not commonly associated with LGMD1B. Using whole genome sequencing (WGS), we identified 21 DNA sequence variants that segregate with the family members possessing more profound dystrophic features and muscle cores. These include a relatively common variant in coiled-coil domain containing protein 78 (CCDC78). This variant was given priority because another mutation in CCDC78 causes autosomal dominant centronuclear myopathy-4, which causes cores in addition to centrally positioned nuclei. Therefore, we analyzed muscle biopsies from family members and discovered that those with both the LMNA mutation and the CCDC78 variant contain muscle cores that accumulated both CCDC78 and RyR1. Muscle cores containing mislocalized CCDC78 and RyR1 were absent in the less profoundly affected family members possessing only the LMNA mutation. Taken together, our findings suggest that a relatively common variant in CCDC78 can impart profound muscle pathology in combination with a LMNA mutation and accounts for variability in skeletal muscle disease phenotypes.

Pleiotropic effects of different exonic nucleotide changes at the same position contribute to hemophilia B phenotypic variation

BACKGROUND

The disease-causing effects of genetic variations often depend on their location within a gene. Exonic changes generally lead to alterations in protein production, secretion, activity, or clearance. However, owing to the overlap between proteins and splicing codes, missense variants can also affect messenger RNA splicing, thus adding a layer of complexity and influencing disease phenotypes.

OBJECTIVES

To extensively characterize a panel of 13 exonic variants in the F9 gene occurring at 6 different factor IX positions and associated with varying severities of hemophilia B (HB).

METHODS

Computational predictions, splicing analysis, and recombinant factor IX assays were exploited to characterize F9 variants.

RESULTS

We demonstrated that 5 (38%) of 13 selected F9 exonic variants have pleiotropic effects. Although bioinformatic approaches accurately classified effects, extensive experimental assays were required to elucidate and deepen the molecular mechanisms underlying the pleiotropic effects. Importantly, their characterization was instrumental in developing tailored RNA therapeutics based on engineered U7 small nuclear RNA to mask cryptic splice sites and compensatory U1 small nuclear RNA to enhance exon definition.

CONCLUSION

Overall, albeit a multitool bioinformatic approach suggested the molecular effects of multiple HB variants, the deep investigation of molecular mechanisms revealed insights into the HB phenotype-genotype relationship, enabling accurate classification of HB variants. Importantly, knowledge of molecular mechanisms allowed the development of tailored RNA therapeutics, which can also be translated to other genetic diseases.

On a kneading theory for gene-splicing

Two well-known facets in protein synthesis in eukaryotic cells are transcription of DNA to pre-RNA in the nucleus and the translation of messenger-RNA (mRNA) to proteins in the cytoplasm. A critical intermediate step is the removal of segments (introns) containing ∼97% of the nucleic-acid sites in pre-RNA and sequential alignment of the retained segments (exons) to form mRNA through a process referred to as splicing. Alternative forms of splicing enrich the proteome while abnormal splicing can enhance the likelihood of a cell developing cancer or other diseases. Mechanisms for splicing and origins of splicing errors are only partially deciphered. Our goal is to determine if rules on splicing can be inferred from data analytics on nucleic-acid sequences. Toward that end, we represent a nucleic-acid site as a point in a plane defined in terms of the anterior and posterior sub-sequences of the site. The "point-set" representation expands analytical approaches, including the use of statistical tools, to characterize genome sequences. It is found that point-sets for exons and introns are visually different, and that the differences can be quantified using a family of generalized moments. We design a machine-learning algorithm that can recognize individual exons or introns with 91% accuracy. Point-set distributions and generalized moments are found to differ between organisms.

DRANetSplicer: A Splice Site Prediction Model Based on Deep Residual Attention Networks

The precise identification of splice sites is essential for unraveling the structure and function of genes, constituting a pivotal step in the gene annotation process. In this study, we developed a novel deep learning model, DRANetSplicer, that integrates residual learning and attention mechanisms for enhanced accuracy in capturing the intricate features of splice sites. We constructed multiple datasets using the most recent versions of genomic data from three different organisms, Oryza sativa japonica, Arabidopsis thaliana and Homo sapiens. This approach allows us to train models with a richer set of high-quality data. DRANetSplicer outperformed benchmark methods on donor and acceptor splice site datasets, achieving an average accuracy of (96.57%, 95.82%) across the three organisms. Comparative analyses with benchmark methods, including SpliceFinder, Splice2Deep, Deep Splicer, EnsembleSplice, and DNABERT, revealed DRANetSplicer's superior predictive performance, resulting in at least a (4.2%, 11.6%) relative reduction in average error rate. We utilized the DRANetSplicer model trained on O. sativa japonica data to predict splice sites in A. thaliana, achieving accuracies for donor and acceptor sites of (94.89%, 94.25%). These results indicate that DRANetSplicer possesses excellent cross-organism predictive capabilities, with its performance in cross-organism predictions even surpassing that of benchmark methods in non-cross-organism predictions. Cross-organism validation showcased DRANetSplicer's excellence in predicting splice sites across similar organisms, supporting its applicability in gene annotation for understudied organisms. We employed multiple methods to visualize the decision-making process of the model. The visualization results indicate that DRANetSplicer can learn and interpret well-known biological features, further validating its overall performance. Our study systematically examined and confirmed the predictive ability of DRANetSplicer from various levels and perspectives, indicating that its practical application in gene annotation is justified.

Circulating cell-free DNA-based biomarkers for prognostication and disease monitoring in adrenocortical carcinoma

OBJECTIVE

Adrenocortical carcinoma (ACC) is a rare aggressive cancer with heterogeneous behaviour. Disease surveillance relies on frequent imaging, which comes with significant radiation exposure. The aim of the study was to investigate the role of circulating cell-free DNA (ccfDNA)-related biomarkers (BMs) for prognostication and monitoring of ACC.

DESIGN AND METHODS

We investigated 34 patients with ACC and 23 healthy subjects (HSs) as controls. Circulating cell-free DNA was extracted by commercial kits and ccfDNA concentrations were quantified by fluorimeter (BM1). Targeted sequencing was performed using a customized panel of 27 ACC-specific genes. Leucocyte DNA was used to discriminate somatic variants (BM2), while tumour DNA was sequenced in 22/34 cases for comparison. Serial ccfDNA samples were collected during follow-up in 19 ACC patients (median period 9 months) and analysed in relationship with standard radiological imaging.

RESULTS

Circulating cell-free DNA concentrations were higher in ACC than HS (mean ± SD, 1.15 ± 1.56 vs 0.05 ± 0.05 ng/µL, P < .0001), 96% of them being above the cut-off of 0.146 ng/µL (mean HS + 2 SD, positive BM1). At ccfDNA sequencing, 47% of ACC showed at least 1 somatic mutation (positive BM2). A combined ccfDNA-BM score was strongly associated with both progression-free and overall survival (hazard ratio [HR] = 2.63; 95% CI, 1.13-6.13; P = .010, and HR = 5.98; 95% CI, 2.29-15.6; P = .0001, respectively). During disease monitoring, positive BM2 showed the best specificity (100%) and sensitivity (67%) to detect ACC recurrence or progress compared with BM1.

CONCLUSION

ccfDNA-related BMs are frequently detected in ACC patients and represent a promising, minimally invasive tool to predict clinical outcome and complement surveillance imaging. Our findings will be validated in a larger cohort of ACCs with long-term follow-up.

Coding and non-coding variants in the ciliopathy gene CFAP410 cause early-onset non-syndromic retinal degeneration

Inherited retinal degenerations are blinding genetic disorders characterized by high genetic and phenotypic heterogeneity. The implementation of next-generation sequencing in routine diagnostics, together with advanced clinical phenotyping including multimodal retinal imaging, have contributed to the increase of reports describing novel genotype-phenotype associations and phenotypic expansions. In this study, we describe sixteen families with early-onset non-syndromic retinal degenerations in which affected probands carried rare bi-allelic variants in CFAP410, a ciliary gene previously associated with syndromic recessive Jeune syndrome. The most common retinal phenotypes were cone-rod and rod-cone dystrophies, but the clinical presentations were unified by their early onset as well as the severe impact on central visual function. Twelve variants were detected (three pathogenic, seven likely pathogenic, two of uncertain significance), eight of which were novel. One deep intronic change, c.373+91A>G, led to the creation of a cryptic splice acceptor site in intron four, followed by the inclusion of a 200- base pair pseudoexon and subsequent premature stop codon formation. To our knowledge this is the first likely pathogenic deep-intronic variant identified in this gene. Meta-analysis of all published and novel CFAP410 variants revealed no clear correlation between the severity of the CFAP410-associated phenotypes and the identified causal variants. This is supported by the fact that the frequently encountered missense variant p.(Arg73Pro), often found in syndromic cases, was also associated with non-syndromic retinal degeneration. This study expands the current knowledge of CFAP410-associated ciliopathy by enriching its mutational landscape and supports its association with non-syndromic retinal degeneration.

Exploring noncoding variants in genetic diseases: from detection to functional insights

Previous studies on genetic diseases predominantly focused on protein-coding variations, overlooking the vast noncoding regions in the human genome. The development of high-throughput sequencing technologies and functional genomics tools has enabled the systematic identification of functional noncoding variants. These variants can impact gene expression, regulation, and chromatin conformation, thereby contributing to disease pathogenesis. Understanding the mechanisms that underlie the impact of noncoding variants on genetic diseases is indispensable for the development of precisely targeted therapies and the implementation of personalized medicine strategies. The intricacies of noncoding regions introduce a multitude of challenges and research opportunities. In this review, we introduce a spectrum of noncoding variants involved in genetic diseases, along with research strategies and advanced technologies for their precise identification and in-depth understanding of the complexity of the noncoding genome. We will delve into the research challenges and propose potential solutions for unraveling the genetic basis of rare and complex diseases.

CRISPR activation to characterize splice-altering variants in easily accessible cells

Genetic variants that affect mRNA splicing are a major cause of hereditary disorders, but the spliceogenicity of variants is challenging to predict. RNA diagnostics of clinically accessible tissues enable rapid functional characterization of splice-altering variants within their natural genetic context. However, this analysis cannot be offered to all individuals as one in five human disease genes are not expressed in easily accessible cell types. To overcome this problem, we have used CRISPR activation (CRISPRa) based on a dCas9-VPR mRNA-based delivery platform to induce expression of the gene of interest in skin fibroblasts from individuals with suspected monogenic disorders. Using this ex vivo splicing assay, we characterized the splicing patterns associated with germline variants in the myelin protein zero gene (MPZ), which is exclusively expressed in Schwann cells of the peripheral nerves, and the spastin gene (SPAST), which is predominantly expressed in the central nervous system. After overnight incubation, CRISPRa strongly upregulated MPZ and SPAST transcription in skin fibroblasts, which enabled splice variant profiling using reverse transcription polymerase chain reaction, next-generation sequencing, and long-read sequencing. Our investigations show proof of principle of a promising genetic diagnostic tool that involves CRISPRa to activate gene expression in easily accessible cells to study the functional impact of genetic variants. The procedure is easy to perform in a diagnostic laboratory with equipment and reagents all readily available.

Whole genome sequencing in clinical practice

Whole genome sequencing (WGS) is becoming the preferred method for molecular genetic diagnosis of rare and unknown diseases and for identification of actionable cancer drivers. Compared to other molecular genetic methods, WGS captures most genomic variation and eliminates the need for sequential genetic testing. Whereas, the laboratory requirements are similar to conventional molecular genetics, the amount of data is large and WGS requires a comprehensive computational and storage infrastructure in order to facilitate data processing within a clinically relevant timeframe. The output of a single WGS analyses is roughly 5 MIO variants and data interpretation involves specialized staff collaborating with the clinical specialists in order to provide standard of care reports. Although the field is continuously refining the standards for variant classification, there are still unresolved issues associated with the clinical application. The review provides an overview of WGS in clinical practice - describing the technology and current applications as well as challenges connected with data processing, interpretation and clinical reporting.

Improved modeling of RNA-binding protein motifs in an interpretable neural model of RNA splicing

Sequence-specific RNA-binding proteins (RBPs) play central roles in splicing decisions. Here, we describe a modular splicing architecture that leverages in vitro-derived RNA affinity models for 79 human RBPs and the annotated human genome to produce improved models of RBP binding and activity. Binding and activity are modeled by separate Motif and Aggregator components that can be mixed and matched, enforcing sparsity to improve interpretability. Training a new Adjusted Motif (AM) architecture on the splicing task not only yields better splicing predictions but also improves prediction of RBP-binding sites in vivo and of splicing activity, assessed using independent data.

Predicting Alu exonization in the human genome with a deep learning model

Alu exonization, or the recruitment of intronic Alu elements into gene sequences, has contributed to functional diversification; however, its extent and the ways in which it influences gene regulation are not fully understood. We developed an unbiased approach to predict Alu exonization events from genomic sequences implemented in a deep learning model, eXAlu, that overcomes the limitations of tissue or condition specificity and the computational burden of RNA-seq analysis. The model captures previously reported characteristics of exonized Alu sequences and can predict sequence elements important for Alu exonization. Using eXAlu, we estimate the number of Alu elements in the human genome undergoing exonization to be between 55-110K, 11-21 fold more than represented in the GENCODE gene database. Using RT-PCR we were able to validate selected predicted Alu exonization events, supporting the accuracy of our method. Lastly, we highlight a potential application of our method to identify polymorphic Alu insertion exonizations in individuals and in the population from whole genome sequencing data.

Identification of Two Novel Pathogenic Variants of the ATM Gene in the Iranian-Azeri Turkish Ethnic Group by Applying Whole Exome Sequencing

Background

The ATM gene encodes a multifunctional kinase involved in important cellular functions, such as checkpoint signaling and apoptosis, in response to DNA damage. Bi-allelic pathogenic variants in this gene cause Ataxia Telangiectasia (AT), while carriers of ATM pathogenic variants are at increased risk of cancer depending on the pathogenicity of the variant they carry. Identifying pathogenic variants can aid in the management of the disease in carriers.

Methods

Whole-exome sequencing (WES) was performed on three unrelated patients from the Iranian-Azeri Turkish ethnic group referred to a genetic center for analysis. WES was also conducted on 400 individuals from the same ethnic group to determine the frequencies of all ATM variants. Blood samples were collected from the patients and their family members for DNA extraction, and PCR-Sanger sequencing was performed to confirm the WES results.

Results

The first proband with AT disease had two novel compound heterozygote variants (c.2639-2A>T, c.8708delC) in the ATM gene revealed by WES analysis, which was potentially/likely pathogenic. The second proband with bi-lateral breast cancer had a homozygous pathogenic variant (c.6067G>A) in the ATM gene identified by WES analysis. The third case with a family history of cancer had a heterozygous synonymous pathogenic variant (c.7788G>A) in the ATM gene found by WES analysis. Sanger sequencing confirmed the WES results, and bioinformatics analysis of the mutated ATM RNA and protein structure added evidence for the potential pathogenicity of the novel variants. WES analysis of the cohort revealed 38 different variants, including a variant (rs1800057, ATM:c.3161C>G, p.P1054R) associated with prostate cancer that had a higher frequency in our cohort.

Conclusion

Genetic analysis of three unrelated families with ATM-related disorders discovered two novel pathogenic variants. A homozygous missense pathogenic variant was identified in a woman with bi-lateral breast cancer, and a synonymous but pathogenic variant was found in a family with a history of different cancers.

Dysregulation of MicroRNAs in Adult Osteogenesis Imperfecta: The miROI Study

As epigenetic regulators of gene expression, circulating micro-RiboNucleic Acids (miRNAs) have been described in several bone diseases as potential prognostic markers. The aim of our study was to identify circulating miRNAs potentially associated with the severity of osteogenesis imperfecta (OI) in three steps. We have screened by RNA sequencing for the miRNAs that were differentially expressed in sera of a small group of OI patients versus controls and then conducted a validation phase by RT-qPCR analysis of sera of a larger patient population. In the first phase of miROI, we found 79 miRNAs that were significantly differentially expressed. We therefore selected 19 of them as the most relevant. In the second phase, we were able to validate the significant overexpression of 8 miRNAs in the larger OI group. Finally, we looked for a relationship between the level of variation of the validated miRNAs and the clinical characteristics of OI. We found a significant difference in the expression of two microRNAs in those patients with dentinogenesis imperfecta. After reviewing the literature, we found 6 of the 8 miRNAs already known to have a direct action on bone homeostasis. Furthermore, the use of a miRNA-gene interaction prediction model revealed a 100% probability of interaction between 2 of the 8 confirmed miRNAs and COL1A1 and/or COL1A2. This is the first study to establish the miRNA signature in OI, showing a significant modification of miRNA expression potentially involved in the regulation of genes involved in the physiopathology of OI. © 2023 The Authors. Journal of Bone and Mineral Research published by Wiley Periodicals LLC on behalf of American Society for Bone and Mineral Research (ASBMR).

Targeted sequencing and in vitro splice assays shed light on ABCA4-associated retinopathies missing heritability

The ABCA4 gene is the most frequently mutated Mendelian retinopathy-associated gene. Biallelic variants lead to a variety of phenotypes, however, for thousands of cases the underlying variants remain unknown. Here, we aim to shed further light on the missing heritability of ABCA4-associated retinopathy by analyzing a large cohort of macular dystrophy probands. A total of 858 probands were collected from 26 centers, of whom 722 carried no or one pathogenic ABCA4 variant, while 136 cases carried two ABCA4 alleles, one of which was a frequent mild variant, suggesting that deep-intronic variants (DIVs) or other cis-modifiers might have been missed. After single molecule molecular inversion probes (smMIPs)-based sequencing of the complete 128-kb ABCA4 locus, the effect of putative splice variants was assessed in vitro by midigene splice assays in HEK293T cells. The breakpoints of copy number variants (CNVs) were determined by junction PCR and Sanger sequencing. ABCA4 sequence analysis solved 207 of 520 (39.8%) naive or unsolved cases and 70 of 202 (34.7%) monoallelic cases, while additional causal variants were identified in 54 of 136 (39.7%) probands carrying two variants. Seven novel DIVs and six novel non-canonical splice site variants were detected in a total of 35 alleles and characterized, including the c.6283-321C>G variant leading to a complex splicing defect. Additionally, four novel CNVs were identified and characterized in five alleles. These results confirm that smMIPs-based sequencing of the complete ABCA4 gene provides a cost-effective method to genetically solve retinopathy cases and that several rare structural and splice altering defects remain undiscovered in Stargardt disease cases.

De novo PHF5A variants are associated with craniofacial abnormalities, developmental delay, and hypospadias

PURPOSE

The SF3B splicing complex is composed of SF3B1-6 and PHF5A. We report a developmental disorder caused by de novo variants in PHF5A.

METHODS

Clinical, genomic, and functional studies using subject-derived fibroblasts and a heterologous cellular system were performed.

RESULTS

We studied 9 subjects with congenital malformations, including preauricular tags and hypospadias, growth abnormalities, and developmental delay who had de novo heterozygous PHF5A variants, including 4 loss-of-function (LOF), 3 missense, 1 splice, and 1 start-loss variant. In subject-derived fibroblasts with PHF5A LOF variants, wild-type and variant PHF5A mRNAs had a 1:1 ratio, and PHF5A mRNA levels were normal. Transcriptome sequencing revealed alternative promoter use and downregulated genes involved in cell-cycle regulation. Subject and control fibroblasts had similar amounts of PHF5A with the predicted wild-type molecular weight and of SF3B1-3 and SF3B6. SF3B complex formation was unaffected in 2 subject cell lines.

CONCLUSION

Our data suggest the existence of feedback mechanisms in fibroblasts with PHF5A LOF variants to maintain normal levels of SF3B components. These compensatory mechanisms in subject fibroblasts with PHF5A or SF3B4 LOF variants suggest disturbed autoregulation of mutated splicing factor genes in specific cell types, that is, neural crest cells, during embryonic development rather than haploinsufficiency as pathomechanism.

Chromosome Microarray Analysis and Exome Sequencing: Implementation in Prenatal Diagnosis of Fetuses with Digestive System Malformations

(1) Purpose: Retrospective back-to-back comparisons were performed to evaluate the accuracy, effectiveness, and incremental yield of chromosome microarray analysis (CMA) and exome sequencing (ES) analysis in fetuses with digestive system malformations (DSMs). (2) Methods: In total, 595 women with fetal DSMs who underwent prenatal diagnosis were enrolled. We analyzed the diagnostic yields of CMA and ES and evaluated pregnancy outcomes. Copy number variants (CNVs) were classified according to the American College of Medical Genetics and Genomics guidelines. (3) Results: Pathogenic CNVs were detected in 11/517 (2.12%) fetuses, and variants of unknown significance (VUS) were identified in 69 (13.35%) fetuses using CMA. ES detected 29 pathogenic/likely pathogenic variants in 23/143 (16.08%) fetuses and 26/143 (18.2%) VUS. In those with other ultrasound abnormalities, the detection rate of multiple system structural malformations was 41.2%, followed by skeletal (33.3%), cardiovascular (25.4%), and central nervous system (18.6%) malformations. Of the 391 surviving children, 40 (10.2%) exhibited varying degrees of mental retardation. (4) Conclusion: A correlation exists between DSMs and chromosomal abnormalities. When combined with other systemic abnormalities, the incidence of chromosomal abnormalities increases significantly. Patients with congenital DSM are at risk of developing neurodevelopmental disorders. Combined CMA and ES detection of fetal DSM has good clinical application potential.

Comparison of disease phenotypes and mechanistic insight on causal variants in patients with DADA2

BACKGROUND

Deficiency of adenosine deaminase 2 (DADA2) results in heterogeneous manifestations including systemic vasculitis and red cell aplasia. The basis of different disease phenotypes remains incompletely defined.

OBJECTIVE

We sought to further delineate disease phenotypes in DADA2 and define the mechanistic basis of ADA2 variants.

METHODS

We analyzed the clinical features and ADA2 variants in 33 patients with DADA2. We compared the transcriptomic profile of 14 patients by bulk RNA sequencing. ADA2 variants were expressed experimentally to determine impact on protein production, trafficking, release, and enzymatic function.

RESULTS

Transcriptomic analysis of PBMCs from DADA2 patients with the vasculitis phenotype or pure red cell aplasia phenotype exhibited similar upregulation of TNF, type I interferon, and type II interferon signaling pathways compared with healthy controls. These pathways were also activated in 3 asymptomatic individuals with DADA2. Analysis of ADA2 variants, including 7 novel variants, showed different mechanisms of functional disruption including (1) unstable transcript leading to RNA degradation; (2) impairment of ADA2 secretion because of retention in the endoplasmic reticulum; (3) normal expression and secretion of ADA2 that lacks enzymatic function; and (4) disruption of the N-terminal signal peptide leading to cytoplasmic localization of unglycosylated protein.

CONCLUSIONS

Transcriptomic signatures of inflammation are observed in patients with different disease phenotypes, including some asymptomatic individuals. Disease-associated ADA2 variants affect protein function by multiple mechanisms, which may contribute to the clinical heterogeneity of DADA2.

Splam: a deep-learning-based splice site predictor that improves spliced alignments

The process of splicing messenger RNA to remove introns plays a central role in creating genes and gene variants. Here we describe Splam, a novel method for predicting splice junctions in DNA based on deep residual convolutional neural networks. Unlike some previous models, Splam looks at a relatively limited window of 400 base pairs flanking each splice site, motivated by the observation that the biological process of splicing relies primarily on signals within this window. Additionally, Splam introduces the idea of training the network on donor and acceptor pairs together, based on the principle that the splicing machinery recognizes both ends of each intron at once. We compare Splam's accuracy to recent state-of-the-art splice site prediction methods, particularly SpliceAI, another method that uses deep neural networks. Our results show that Splam is consistently more accurate than SpliceAI, with an overall accuracy of 96% at predicting human splice junctions. Splam generalizes even to non-human species, including distant ones like the flowering plant Arabidopsis thaliana. Finally, we demonstrate the use of Splam on a novel application: processing the spliced alignments of RNA-seq data to identify and eliminate errors. We show that when used in this manner, Splam yields substantial improvements in the accuracy of downstream transcriptome analysis of both poly(A) and ribo-depleted RNA-seq libraries. Overall, Splam offers a faster and more accurate approach to detecting splice junctions, while also providing a reliable and efficient solution for cleaning up erroneous spliced alignments.

An automated framework for evaluation of deep learning models for splice site predictions

A novel framework for the automated evaluation of various deep learning-based splice site detectors is presented. The framework eliminates time-consuming development and experimenting activities for different codebases, architectures, and configurations to obtain the best models for a given RNA splice site dataset. RNA splicing is a cellular process in which pre-mRNAs are processed into mature mRNAs and used to produce multiple mRNA transcripts from a single gene sequence. Since the advancement of sequencing technologies, many splice site variants have been identified and associated with the diseases. So, RNA splice site prediction is essential for gene finding, genome annotation, disease-causing variants, and identification of potential biomarkers. Recently, deep learning models performed highly accurately for classifying genomic signals. Convolutional Neural Network (CNN), Long Short-Term Memory (LSTM) and its bidirectional version (BLSTM), Gated Recurrent Unit (GRU), and its bidirectional version (BGRU) are promising models. During genomic data analysis, CNN's locality feature helps where each nucleotide correlates with other bases in its vicinity. In contrast, BLSTM can be trained bidirectionally, allowing sequential data to be processed from forward and reverse directions. Therefore, it can process 1-D encoded genomic data effectively. Even though both methods have been used in the literature, a performance comparison was missing. To compare selected models under similar conditions, we have created a blueprint for a series of networks with five different levels. As a case study, we compared CNN and BLSTM models' learning capabilities as building blocks for RNA splice site prediction in two different datasets. Overall, CNN performed better with [Formula: see text] accuracy ([Formula: see text] improvement), [Formula: see text] F1 score ([Formula: see text] improvement), and [Formula: see text] AUC-PR ([Formula: see text] improvement) in human splice site prediction. Likewise, an outperforming performance with [Formula: see text] accuracy ([Formula: see text] improvement), [Formula: see text] F1 score ([Formula: see text] improvement), and [Formula: see text] AUC-PR ([Formula: see text] improvement) is achieved in C. elegans splice site prediction. Overall, our results showed that CNN learns faster than BLSTM and BGRU. Moreover, CNN performs better at extracting sequence patterns than BLSTM and BGRU. To our knowledge, no other framework is developed explicitly for evaluating splice detection models to decide the best possible model in an automated manner. So, the proposed framework and the blueprint would help selecting different deep learning models, such as CNN vs. BLSTM and BGRU, for splice site analysis or similar classification tasks and in different problems.

Functional characterization of human genomic variation linked to polygenic diseases

The burden of human disease lies predominantly in polygenic diseases. Since the early 2000s, genome-wide association studies (GWAS) have identified genetic variants and loci associated with complex traits. These have ranged from variants in coding sequences to mutations in regulatory regions, such as promoters and enhancers, as well as mutations affecting mediators of mRNA stability and other downstream regulators, such as 5' and 3'-untranslated regions (UTRs), long noncoding RNA (lncRNA), and miRNA. Recent research advances in genetics have utilized a combination of computational techniques, high-throughput in vitro and in vivo screening modalities, and precise genome editing to impute the function of diverse classes of genetic variants identified through GWAS. In this review, we highlight the vastness of genomic variants associated with polygenic disease risk and address recent advances in how genetic tools can be used to functionally characterize them.

Discovering cryptic splice mutations in cancers via a deep neural network framework

Somatic mutations can disrupt splicing regulatory elements and have dramatic effects on cancer genes, yet the functional consequences of mutations located in extended splice regions is difficult to predict. Here, we use a deep neural network (SpliceAI) to characterize the landscape of splice-altering mutations in cancer. In our in-house series of 401 liver cancers, SpliceAI uncovers 1244 cryptic splice mutations, located outside essential splice sites, that validate at a high rate (66%) in matched RNA-seq data. We then extend the analysis to a large pan-cancer cohort of 17 714 tumors, revealing >100 000 cryptic splice mutations. Taking into account these mutations increases the power of driver gene discovery, revealing 126 new candidate driver genes. It also reveals new driver mutations in known cancer genes, doubling the frequency of splice alterations in tumor suppressor genes. Mutational signature analysis suggests mutational processes that could give rise preferentially to splice mutations in each cancer type, with an enrichment of signatures related to clock-like processes and DNA repair deficiency. Altogether, this work sheds light on the causes and impact of cryptic splice mutations in cancer, and highlights the power of deep learning approaches to better annotate the functional consequences of mutations in oncology.

CNNSplice: Robust models for splice site prediction using convolutional neural networks

The identification of splice site, or segments of an RNA gene where noncoding and coding sequences are connected in the 5' and 3' directions, is an essential post-transcriptional step for the annotation of functional genes and is required for the study and analysis of biological function in eukaryotic organisms through protein production and gene expression. Splice site detection tools have been proposed for this purpose; however, the models of these tools have a specific use case and are inefficiently or typically untransferable between organisms. Here, we present CNNSplice, a set of deep convolutional neural network models for splice site prediction. Using the five-fold cross-validation model selection technique, we explore several models based on typical machine learning applications and propose five high-performing models to efficiently predict the true and false SS in balanced and imbalanced datasets. Our evaluation results indicate that CNNSplice's models achieve a better performance compared with existing methods across five organisms' datasets. In addition, our generality test shows CNNSplice's model ability to predict and annotate splice sites in new or poorly trained genome datasets indicating a broad application spectrum. CNNSplice demonstrates improved model prediction, interpretability, and generalizability on genomic datasets compared to existing splice site prediction tools. We have developed a web server for the CNNSplice algorithm which can be publicly accessed here: http://www.cnnsplice.online.

In silico methods for predicting functional synonymous variants

Single nucleotide variants (SNVs) contribute to human genomic diversity. Synonymous SNVs are previously considered to be "silent," but mounting evidence has revealed that these variants can cause RNA and protein changes and are implicated in over 85 human diseases and cancers. Recent improvements in computational platforms have led to the development of numerous machine-learning tools, which can be used to advance synonymous SNV research. In this review, we discuss tools that should be used to investigate synonymous variants. We provide supportive examples from seminal studies that demonstrate how these tools have driven new discoveries of functional synonymous SNVs.

Aberrant splicing prediction across human tissues

Aberrant splicing is a major cause of genetic disorders but its direct detection in transcriptomes is limited to clinically accessible tissues such as skin or body fluids. While DNA-based machine learning models can prioritize rare variants for affecting splicing, their performance in predicting tissue-specific aberrant splicing remains unassessed. Here we generated an aberrant splicing benchmark dataset, spanning over 8.8 million rare variants in 49 human tissues from the Genotype-Tissue Expression (GTEx) dataset. At 20% recall, state-of-the-art DNA-based models achieve maximum 12% precision. By mapping and quantifying tissue-specific splice site usage transcriptome-wide and modeling isoform competition, we increased precision by threefold at the same recall. Integrating RNA-sequencing data of clinically accessible tissues into our model, AbSplice, brought precision to 60%. These results, replicated in two independent cohorts, substantially contribute to noncoding loss-of-function variant identification and to genetic diagnostics design and analytics.

Analysis of 200 unrelated individuals with a constitutional NF1 deep intronic pathogenic variant reveals that variants flanking the alternatively spliced NF1 exon 31 [23a] cause a classical neurofibromatosis type 1 phenotype while altering predominantly NF1 isoform type II

Neurofibromatosis type 1 results from loss-of-function NF1 pathogenic variants (PVs). Up to 30% of all NF1 PVs disrupt mRNA splicing, including deep intronic variants. Here, we retrospectively investigated the spectrum of NF1 deep intronic PVs in a cohort of 8,090 unrelated individuals from the University of Alabama at Birmingham (UAB) dataset with a molecularly confirmed neurofibromatosis type 1. All variants were identified through their effect on the NF1 transcript, followed by variant characterization at the DNA-level. A total of 68 distinct variants, which were ≥ 20 nucleotides away from the closest exon-intron junction, were identified in 2.5% unrelated individuals with NF1 (200/8,090). Nine different pathogenic splice variants, identified in 20 probands, led to exonization of different parts of intron 30 [23.2] or 31 [23a]. The two major NF1 transcript isoforms, distinguished by the absence (type I) or presence (type II) of the alternatively spliced cassette exon 31 [23a], are equally expressed in blood in control individuals without NF1 or NF1-affected individuals carrying their PV not in the introns flanking exon 31 [23a]. By fragment and cloning analysis we demonstrated that the exonization of intron 31 [23a] sequences due to deep intronic PV predominantly affects the NF1 isoform II. Seven additional (likely) pathogenic NF1 deep intronic variants not observed in the UAB dataset were found by classification of 36 variants identified by a literature search. Hence, the unique list of these 75 deep intronic (likely) PVs should be included in any comprehensive NF1 testing strategy.

Functional Analysis of a Novel, Non-Canonical RPGR Splice Variant Causing X-Linked Retinitis Pigmentosa

X-linked retinitis pigmentosa (XLRP) caused by mutations in the RPGR gene is one of the most severe forms of RP due to its early onset and intractable progression. Most cases have been associated with genetic variants within the purine-rich exon ORF15 region of this gene. RPGR retinal gene therapy is currently being investigated in several clinical trials. Therefore, it is crucial to report and functionally characterize (all novel) potentially pathogenic DNA sequence variants. Whole-exome sequencing (WES) was performed for the index patient. The splicing effects of a non-canonical splice variant were tested on cDNA from whole blood and a minigene assay. WES revealed a rare, non-canonical splice site variant predicted to disrupt the wildtype splice acceptor and create a novel acceptor site 8 nucleotides upstream of RPGR exon 12. Reverse-transcription PCR analyses confirmed the disruption of the correct splicing pattern, leading to the insertion of eight additional nucleotides in the variant transcript. Transcript analyses with minigene assays and cDNA from peripheral blood are useful tools for the characterization of splicing defects due to variants in the RPGR and may increase the diagnostic yield in RP. The functional analysis of non-canonical splice variants is required to classify those variants as pathogenic according to the ACMG's criteria.

Regulation of pre-mRNA splicing: roles in physiology and disease, and therapeutic prospects

The removal of introns from mRNA precursors and its regulation by alternative splicing are key for eukaryotic gene expression and cellular function, as evidenced by the numerous pathologies induced or modified by splicing alterations. Major recent advances have been made in understanding the structures and functions of the splicing machinery, in the description and classification of physiological and pathological isoforms and in the development of the first therapies for genetic diseases based on modulation of splicing. Here, we review this progress and discuss important remaining challenges, including predicting splice sites from genomic sequences, understanding the variety of molecular mechanisms and logic of splicing regulation, and harnessing this knowledge for probing gene function and disease aetiology and for the design of novel therapeutic approaches.

Improving Hereditary Hemorrhagic Telangiectasia Molecular Diagnosis: A Referral Center Experience

BACKGROUND

Hereditary hemorrhagic telangiectasia (HHT) is a rare vascular disease inherited in an autosomal dominant manner. Disease-causing variants in endoglin (ENG) and activin A receptor type II-like 1 (ACVRL1) genes are detected in more than 90% of the patients undergoing molecular testing. The identification of variants of unknown significance is often seen as a challenge in clinical practice that makes family screening and genetic counseling difficult. Here, we show that the implementation of cDNA analysis to assess the effect of splice site variants on mRNA splicing is a powerful tool.

METHODS

Gene panel sequencing of genes associated with HHT and other arteriovenous malformation-related syndromes was performed. To evaluate the effect of the splice site variants, cDNA analysis of ENG and ACVRL1 genes was carried out.

RESULTS

three novel splice site variants were identified in ENG (c.68-2A > T and c.1311+4_1311+8del) and ACVLR1 (c.526-6C > G) genes correspondingly in three individuals with HHT that met ≥ 3 Curaçao criteria. All three variants led to an aberrant splicing inducing exon skipping (ENG:c.68-2A > T and ACVRL1:c.526-6C > G) or intron retention (ENG:c.1311+4_1311+8del) allowing the confirmation of the predicted effect on splicing and the reclassification from unknown significance to pathogenic/likely pathogenic of two of them.

CONCLUSIONS

RNA analysis should be performed to assess and/or confirm the impact of variants on splicing. The molecular diagnosis of HHT patients is crucial to allow family screening and accurate genetic counseling. A multidisciplinary approach including clinicians and geneticists is crucial when dealing with patients with rare diseases.

Unexpected Findings in Hereditary Breast and Ovarian Cancer Syndrome: Low-Level Constitutional Mosaicism in BRCA2

Hereditary breast and ovarian cancer syndrome (HBOC) is a clinical entity characterized by an increased risk of developing breast and ovarian cancer. The genetic diagnosis is based on the identification of heterozygous germinal variants in HBOC susceptibility genes. However, it has recently been described that constitutional mosaic variants can contribute to the aetiology of HBOC. In constitutional mosaicism, individuals have at least two genotypically distinct populations of cells that arise from an early post-zygote event. The mutational event occurs early enough in development to affect several tissues. It is detected in germinal genetic studies as low variant allele frequency (VAF) variants (<30%) that are generally overlooked during the prioritization process. Constitutional mosaic variants can affect both somatic and germinal cells, and thus can be passed to the offspring and have important consequences for genetic counselling. In this work, we report the c.9648+1G>A mosaic variant in the BRCA2 gene and propose a diagnostic algorithm to deal with potential mosaic findings identified by Next Generation Sequencing (NGS).

Cost-effective sequence analysis of 113 genes in 1,192 probands with retinitis pigmentosa and Leber congenital amaurosis

Introduction: Retinitis pigmentosa (RP) and Leber congenital amaurosis (LCA) are two groups of inherited retinal diseases (IRDs) where the rod photoreceptors degenerate followed by the cone photoreceptors of the retina. A genetic diagnosis for IRDs is challenging since >280 genes are associated with these conditions. While whole exome sequencing (WES) is commonly used by diagnostic facilities, the costs and required infrastructure prevent its global applicability. Previous studies have shown the cost-effectiveness of sequence analysis using single molecule Molecular Inversion Probes (smMIPs) in a cohort of patients diagnosed with Stargardt disease and other maculopathies. Methods: Here, we introduce a smMIPs panel that targets the exons and splice sites of all currently known genes associated with RP and LCA, the entire RPE65 gene, known causative deep-intronic variants leading to pseudo-exons, and part of the RP17 region associated with autosomal dominant RP, by using a total of 16,812 smMIPs. The RP-LCA smMIPs panel was used to screen 1,192 probands from an international cohort of predominantly RP and LCA cases. Results and discussion: After genetic analysis, a diagnostic yield of 56% was obtained which is on par with results from WES analysis. The effectiveness and the reduced costs compared to WES renders the RP-LCA smMIPs panel a competitive approach to provide IRD patients with a genetic diagnosis, especially in countries with restricted access to genetic testing.

Optimized Detection of Unknown MET Exon 14 Skipping Mutations in Routine Testing for Patients With Non-Small-Cell Lung Cancer

PURPOSE

MET exon 14 (METex14) skipping is an actionable biomarker in non-small-cell lung cancer. However, MET variants are highly complex and diverse, and not all variants lead to exon 14 skipping. Assessing the skipping effect of unknown variants is still a key issue in molecular diagnosis.

MATERIALS AND METHODS

We retrospectively collected MET variants around exon 14 from 4,233 patients with non-small-cell lung cancer who underwent next-generation sequencing testing using DNA, as well as two published data sets.

RESULTS

Among the 4,233 patients, 44 unique variants including 29 novel variants (65.9%) were discovered from 53 patients. Notably, 31 samples (58.5%) failed RNA verification. Using RNA verification, nine novel skipping variants and five nonskipping variants were confirmed. We further used SpliceAI with the delta score cutoff of 0.315 to aid the classification of novel variants (sensitivity = 98.88% and specificity = 100%). When applied to the reported variants, we also found three wrongly classified nonskipping variants. Finally, an optimized knowledge-based interpretation procedure for clinical routine was built according to the mutation type and location, and five more skipping mutations from the 13 unknown variants were determined, which improved the population determination rate to 0.92%.

CONCLUSION

This study discovered more METex14 skipping variants and optimized an innovative approach that could be adapted for the interpretation of infrequent or novel METex14 variants timely without experimental validation.

Systematic optimization of Cas12a base editors in wheat and maize using the ITER platform

BACKGROUND

Testing an ever-increasing number of CRISPR components is challenging when developing new genome engineering tools. Plant biotechnology has few high-throughput options to perform iterative design-build-test-learn cycles of gene-editing reagents. To bridge this gap, we develop ITER (Iterative Testing of Editing Reagents) based on 96-well arrayed protoplast transfections and high-content imaging.

RESULTS

We validate ITER in wheat and maize protoplasts using Cas9 cytosine and adenine base editors (ABEs), allowing one optimization cycle - from design to results - within 3 weeks. Given that previous LbCas12a-ABEs have low or no activity in plants, we use ITER to develop an optimized LbCas12a-ABE. We show that sequential improvement of five components - NLS, crRNA, LbCas12a, adenine deaminase, and linker - leads to a remarkable increase in activity from almost undetectable levels to 40% on an extrachromosomal GFP reporter. We confirm the activity of LbCas12a-ABE at endogenous targets in protoplasts and obtain base-edited plants in up to 55% of stable wheat transformants and the edits are transmitted to T1 progeny. We leverage these improvements to develop a highly mutagenic LbCas12a nuclease and a LbCas12a-CBE demonstrating that the optimizations can be broadly applied to the Cas12a toolbox.

CONCLUSION

Our data show that ITER is a sensitive, versatile, and high-throughput platform that can be harnessed to accelerate the development of genome editing technologies in plants. We use ITER to create an efficient Cas12a-ABE by iteratively testing a large panel of vector components. ITER will likely be useful to create and optimize genome editing reagents in a wide range of plant species.

Diagnostic Yield of Genetic Testing for Ocular and Oculocutaneous Albinism in a Diverse United States Pediatric Population

The diagnostic yield of genetic testing for ocular/oculocutaneous albinism (OA/OCA) in a diverse pediatric population in the United States (U.S.) is unclear. Phenotypes of 53 patients who presented between 2006-2022 with OA/OCA were retrospectively correlated with genetic testing results. Genetic diagnostic yield was defined as detection of pathogenic/likely pathogenic variant(s) matching the anticipated inheritance for that gene-disease relationship. Variant reclassifications of those with variants of uncertain significance (VUS) and without positive diagnostic yield were completed. Overall initial genetic diagnostic yield of OA/OCA was 66%. There was no significant difference (p = 0.59) between race and ethnicities (Black (78%), White (59%), Hispanic/Latino (64%)); however, the diagnostic yield of OA (33%) was significantly lower (p = 0.007) than OCA (76%). Causative variants in OCA2 (28%) and TYR (20%) were most common. Further, Hermansky-Pudlak syndrome variants were identified in 9% of patients. Re-classification of VUS in non-diagnostic cases resulted in genetic diagnoses for 29% of individuals and increased overall diagnostic yield to 70% of all subjects. There is a high diagnostic yield of genetic testing of patients overall with OA/OCA in a diverse U.S. based pediatric population. Presence or absence of cutaneous involvement of albinism significantly affects genetic diagnostic yield.

Interplay of heavy chain introns influences efficient transcript splicing and affects product quality of recombinant biotherapeutic antibodies from CHO cells

Introns are included in genes encoding therapeutic proteins for their well-documented function of boosting expression. However, mis-splicing of introns in recombinant immunoglobulin (IgG) heavy chain (HC) transcripts can produce amino acid sequence product variants. These variants can affect product quality; therefore, purification process optimization may be needed to remove them, or if they cannot be removed, then in-depth characterization must be carried out to understand their effects on biological activity. In this study, HC transgene engineering approaches were investigated and were successful in significantly reducing the previously identified IgG HC splice variants to <0.5%. Subsequently, a comprehensive evaluation was conducted to understand the influence of the different introns in the HC genes on the expression of recombinant biotherapeutic antibodies. The data revealed an unexpected cooperation between specific introns for efficient splicing, where intron retention led to significant reductions in IgG expression of up to 75% for some intron combinations. Furthermore, it was shown that HC introns could be fully removed without significantly affecting productivity. This work paves the way for future biotherapeutic antibody transgene design with regard to inclusion of HC introns. By removing unnecessary introns, transgene mRNA transcript will no longer be mis-spliced, thereby eliminating HC splice variants and improving antibody product quality.

Functional Assays Reclassify Suspected Splice-Altering Variants of Uncertain Significance in Mendelian Channelopathies

BACKGROUND

Rare protein-altering variants in SCN5A, KCNQ1, and KCNH2 are major causes of Brugada syndrome and the congenital long QT syndrome. While splice-altering variants lying outside 2-bp canonical splice sites can cause these diseases, their role remains poorly described. We implemented 2 functional assays to assess 12 recently reported putative splice-altering variants of uncertain significance and 1 likely pathogenic variant without functional data observed in Brugada syndrome and long QT syndrome probands.

METHODS

We deployed minigene assays to assess the splicing consequences of 10 variants. Three variants incompatible with the minigene approach were introduced into control induced pluripotent stem cells by CRISPR genome editing. We differentiated cells into induced pluripotent stem cell-derived cardiomyocytes and studied splicing outcomes by reverse transcription-polymerase chain reaction. We used the American College of Medical Genetics and Genomics functional assay criteria (PS3/BS3) to reclassify variants.

RESULTS

We identified aberrant splicing, with presumed disruption of protein sequence, in 8/10 variants studied using the minigene assay and 1/3 studied in induced pluripotent stem cell-derived cardiomyocytes. We reclassified 8 variants of uncertain significance to likely pathogenic, 1 variant of uncertain significance to likely benign, and 1 likely pathogenic variant to pathogenic.

CONCLUSIONS

Functional assays reclassified splice-altering variants outside canonical splice sites in Brugada Syndrome- and long QT syndrome-associated genes.

Deep intronic NIPBL de novo mutations and differential diagnoses revealed by whole genome and RNA sequencing in Cornelia de Lange syndrome patients

Cornelia de Lange syndrome (CdLS; MIM# 122470) is a rare developmental disorder. Pathogenic variants in 5 genes explain approximately 50% cases, leaving the other 50% unsolved. We performed whole genome sequencing (WGS) ± RNA sequencing (RNA-seq) in 5 unsolved trios fulfilling the following criteria: (i) clinical diagnosis of classic CdLS, (ii) negative gene panel sequencing from blood and saliva-isolated DNA, (iii) unaffected parents' DNA samples available and (iv) proband's blood-isolated RNA available. A pathogenic de novo mutation (DNM) was observed in a CdLS differential diagnosis gene in 3/5 patients, namely POU3F3, SPEN, and TAF1. In the other two, we identified two distinct deep intronic DNM in NIPBL predicted to create a novel splice site. RT-PCRs and RNA-Seq showed aberrant transcripts leading to the creation of a novel frameshift exon. Our findings suggest the relevance of WGS in unsolved suspected CdLS cases and that deep intronic variants may account for a proportion of them.

Using single molecule Molecular Inversion Probes as a cost-effective, high-throughput sequencing approach to target all genes and loci associated with macular diseases

Macular degenerations (MDs) are a subgroup of retinal disorders characterized by central vision loss. Knowledge is still lacking on the extent of genetic and nongenetic factors influencing inherited MD (iMD) and age-related MD (AMD) expression. Single molecule Molecular Inversion Probes (smMIPs) have proven effective in sequencing the ABCA4 gene in patients with Stargardt disease to identify associated coding and noncoding variation, however many MD patients still remain genetically unexplained. We hypothesized that the missing heritability of MDs may be revealed by smMIPs-based sequencing of all MD-associated genes and risk factors. Using 17,394 smMIPs, we sequenced the coding regions of 105 iMD and AMD-associated genes and noncoding or regulatory loci, known pseudo-exons, and the mitochondrial genome in two test cohorts that were previously screened for variants in ABCA4. Following detailed sequencing analysis of 110 probands, a diagnostic yield of 38% was observed. This established an ''MD-smMIPs panel," enabling a genotype-first approach in a high-throughput and cost-effective manner, whilst achieving uniform and high coverage across targets. Further analysis will identify known and novel variants in MD-associated genes to offer an accurate clinical diagnosis to patients. Furthermore, this will reveal new genetic associations for MD and potential genetic overlaps between iMD and AMD.

SPiP: Splicing Prediction Pipeline, a machine learning tool for massive detection of exonic and intronic variant effects on mRNA splicing

Modeling splicing is essential for tackling the challenge of variant interpretation as each nucleotide variation can be pathogenic by affecting pre-mRNA splicing via disruption/creation of splicing motifs such as 5'/3' splice sites, branch sites, or splicing regulatory elements. Unfortunately, most in silico tools focus on a specific type of splicing motif, which is why we developed the Splicing Prediction Pipeline (SPiP) to perform, in one single bioinformatic analysis based on a machine learning approach, a comprehensive assessment of the variant effect on different splicing motifs. We gathered a curated set of 4616 variants scattered all along the sequence of 227 genes, with their corresponding splicing studies. The Bayesian analysis provided us with the number of control variants, that is, variants without impact on splicing, to mimic the deluge of variants from high-throughput sequencing data. Results show that SPiP can deal with the diversity of splicing alterations, with 83.13% sensitivity and 99% specificity to detect spliceogenic variants. Overall performance as measured by area under the receiving operator curve was 0.986, better than SpliceAI and SQUIRLS (0.965 and 0.766) for the same data set. SPiP lends itself to a unique suite for comprehensive prediction of spliceogenicity in the genomic medicine era. SPiP is available at: https://sourceforge.net/projects/splicing-prediction-pipeline/.

Combining genetic constraint with predictions of alternative splicing to prioritize deleterious splicing in rare disease studies

Background

Despite numerous molecular and computational advances, roughly half of patients with a rare disease remain undiagnosed after exome or genome sequencing. A particularly challenging barrier to diagnosis is identifying variants that cause deleterious alternative splicing at intronic or exonic loci outside of canonical donor or acceptor splice sites.

Results

Several existing tools predict the likelihood that a genetic variant causes alternative splicing. We sought to extend such methods by developing a new metric that aids in discerning whether a genetic variant leads to deleterious alternative splicing. Our metric combines genetic variation in the Genome Aggregate Database with alternative splicing predictions from SpliceAI to compare observed and expected levels of splice-altering genetic variation. We infer genic regions with significantly less splice-altering variation than expected to be constrained. The resulting model of regional splicing constraint captures differential splicing constraint across gene and exon categories, and the most constrained genic regions are enriched for pathogenic splice-altering variants. Building from this model, we developed ConSpliceML. This ensemble machine learning approach combines regional splicing constraint with multiple per-nucleotide alternative splicing scores to guide the prediction of deleterious splicing variants in protein-coding genes. ConSpliceML more accurately distinguishes deleterious and benign splicing variants than state-of-the-art splicing prediction methods, especially in “cryptic” splicing regions beyond canonical donor or acceptor splice sites.

Conclusion

Integrating a model of genetic constraint with annotations from existing alternative splicing tools allows ConSpliceML to prioritize potentially deleterious splice-altering variants in studies of rare human diseases.

Assessment of an automated approach for variant interpretation in screening for monogenic disorders: A single-center study

BACKGROUND

Automation has been introduced into variant interpretation, but it is not known how automated variant interpretation performs on a stand-alone basis. The purpose of this study was to evaluate a fully automated computerized approach.

METHOD

We reviewed all variants encountered in a set of carrier screening panels over a 1-year interval. Observed variants with high-confidence ClinVar interpretations were included in the analysis; those without high-confidence ClinVar entries were excluded.

RESULTS

Discrepancy rates between automated interpretations and high-confidence ClinVar entries were analyzed. Of the variants interpreted as positive (likely pathogenic or pathogenic) based on ClinVar information, 22.6% were classified as negative (variants of uncertain significance, likely benign or benign) variants by the automated method. Of the ClinVar negative variants, 1.7% were classified as positive by the automated software. On a per-case basis, which accounts for variant frequency, 63.4% of cases with a ClinVar high-confidence positive variant were classified as negative by the automated method.

CONCLUSION

While automation in genetic variant interpretation holds promise, there is still a need for manual review of the output. Additional validation of automated variant interpretation methods should be conducted.

Mechanism and modeling of human disease-associated near-exon intronic variants that perturb RNA splicing

It is estimated that 10%-30% of disease-associated genetic variants affect splicing. Splicing variants may generate deleteriously altered gene product and are potential therapeutic targets. However, systematic diagnosis or prediction of splicing variants is yet to be established, especially for the near-exon intronic splice region. The major challenge lies in the redundant and ill-defined branch sites and other splicing motifs therein. Here, we carried out unbiased massively parallel splicing assays on 5,307 disease-associated variants that overlapped with branch sites and collected 5,884 variants across the 5' splice region. We found that strong splice sites and exonic features preserve splicing from intronic sequence variation. Whereas the splice-altering mechanism of the 3' intronic variants is complex, that of the 5' is mainly splice-site destruction. Statistical learning combined with these molecular features allows precise prediction of altered splicing from an intronic variant. This statistical model provides the identity and ranking of biological features that determine splicing, which serves as transferable knowledge and out-performs the benchmarking predictive tool. Moreover, we demonstrated that intronic splicing variants may associate with disease risks in the human population. Our study elucidates the mechanism of splicing response of intronic variants, which classify disease-associated splicing variants for the promise of precision medicine.

Monoclonal Antibody Sequence Variants Disguised as Fragments: Identification, Characterization, and Their Removal by Purification Process Optimization

During early stage development of a therapeutic IgG1 monoclonal antibody, high levels of low molecular weight (LMW) peaks were observed by high performance size-exclusion chromatography and capillary electrophoresis. Further characterization of the LMW peak enriched HPSEC fractions using reversed phase liquid chromatography coupled to mass spectrometry showed these LMW species were 47 kDa and 50 kDa in size. However, the measured masses could not be matched to any fragments resulting from peptide bond hydrolysis. To identify these unknown LMW species, molecular characterization methods were employed, including high-throughput sequencing of RNA. Transcriptomic analysis revealed the LMW species were generated by mis-splicing events in the heavy chain transcript, which produced truncated heavy chain products that assembled with the light chain to mimic the appearance of fragments identified by routine purity assays. In an effort to improve product quality, an optimized purification process was developed. Characterization of the process intermediates confirmed removal of both LMW species by the optimized process. Our study demonstrates that deep-dive analytical characterization of biotherapeutics is critical to ensure product quality and inform process development. Transcriptomic analysis tools can help identify the cause of unknown species, and plays a key role in product and process characterization.

Comparison of In Silico Tools for Splice-Altering Variant Prediction Using Established Spliceogenic Variants: An End-User’s Point of View

Assessing the impact of variants of unknown significance on splicing has become a critical issue and a bottleneck, especially with the widespread implementation of whole-genome or exome sequencing. Although multiple in silico tools are available, the interpretation and application of these tools are difficult and practical guidelines are still lacking. A streamlined decision-making process can facilitate the downstream RNA analysis in a more efficient manner. Therefore, we evaluated the performance of 8 in silico tools (Splice Site Finder, MaxEntScan, Splice-site prediction by neural network, GeneSplicer, Human Splicing Finder, SpliceAI, Splicing Predictions in Consensus Elements, and SpliceRover) using 114 NF1 spliceogenic variants, experimentally validated at the mRNA level. The change in the predicted score incurred by the variant of the nearest wild-type splice site was analyzed, and for type II, III, and IV splice variants, the change in the prediction score of de novo or cryptic splice site was also analyzed. SpliceAI and SpliceRover, tools based on deep learning, outperformed all other tools, with AUCs of 0.972 and 0.924, respectively. For de novo and cryptic splice sites, SpliceAI outperformed all other tools and showed a sensitivity of 95.7% at an optimal cut-off of 0.02 score change. Our results show that deep learning algorithms, especially those of SpliceAI, are validated at a significantly higher rate than other in silico tools for clinically relevant NF1 variants. This suggests that deep learning algorithms outperform traditional probabilistic approaches and classical machine learning tools in predicting the de novo and cryptic splice sites.

EnsembleSplice: ensemble deep learning model for splice site prediction

BACKGROUND

Identifying splice site regions is an important step in the genomic DNA sequencing pipelines of biomedical and pharmaceutical research. Within this research purview, efficient and accurate splice site detection is highly desirable, and a variety of computational models have been developed toward this end. Neural network architectures have recently been shown to outperform classical machine learning approaches for the task of splice site prediction. Despite these advances, there is still considerable potential for improvement, especially regarding model prediction accuracy, and error rate.

RESULTS

Given these deficits, we propose EnsembleSplice, an ensemble learning architecture made up of four (4) distinct convolutional neural networks (CNN) model architecture combination that outperform existing splice site detection methods in the experimental evaluation metrics considered including the accuracies and error rates. We trained and tested a variety of ensembles made up of CNNs and DNNs using the five-fold cross-validation method to identify the model that performed the best across the evaluation and diversity metrics. As a result, we developed our diverse and highly effective splice site (SS) detection model, which we evaluated using two (2) genomic Homo sapiens datasets and the Arabidopsis thaliana dataset. The results showed that for of the Homo sapiens EnsembleSplice achieved accuracies of 94.16% for one of the acceptor splice sites and 95.97% for donor splice sites, with an error rate for the same Homo sapiens dataset, 4.03% for the donor splice sites and 5.84% for the acceptor splice sites datasets.

CONCLUSIONS

Our five-fold cross validation ensured the prediction accuracy of our models are consistent. For reproducibility, all the datasets used, models generated, and results in our work are publicly available in our GitHub repository here: https://github.com/OluwadareLab/EnsembleSplice.

Eosinophilic Infiltration of the Sino-Atrial Node in Sudden Cardiac Death Caused by Long QT Syndrome

Sudden death is defined as the unexpected death of a healthy person that occurs within the first hour of the onset of symptoms or within 24 h of the victim being last seen alive. In some of these cases, rare deleterious variants of genes associated with inherited cardiac disorders can provide a highly probable explanation for the fatal event. We report the case of a 21-year-old obese woman who lost consciousness suddenly in a public place and was pronounced dead after hospital admission. Clinical autopsy showed an inconclusive gross examination, while in the histopathological analysis an eosinophilic inflammatory focus and interstitial fibrosis in the sino-atrial node were found. Molecular autopsy revealed an intronic variant in the KCNQ1 gene (c.683 + 5G > A), classified as likely pathogenic for long QT syndrome according to the guidelines provided by the American College of Medical Genetics and Genomics. Therefore, there were many anomalies that could have played a role in the causation of the sudden death, such as the extreme obesity, the cardiac anomalies and the KNCQ1 variant. This case depicts the difficult interpretation of rare cardiac structural abnormalities in subjects carrying rare variants responsible for inherited arrhythmic disorders and the challenge for the forensic pathologist to make causal inferences in the determinism of the unexpected decease.

Targeted Next Generation Sequencing Revealed Novel Variants in the PKD1 and PKD2 Genes of Iranian Patients with Autosomal Dominant Polycystic Kidney Disease

BACKGROUND

Autosomal dominant polycystic kidney disease (ADPKD), one of the common inherited disorders in humans, is characterized by the development and enlargement of renal cysts, often leading to end-stage renal disease (ESRD). In this study, Iranian ADPKD families were subjected to high-throughput DNA sequencing to find potential causative variants facilitating the way toward risk assessment and targeted therapy.

METHODS

Our protocol was based on the targeted next generation sequencing (NGS) panel previously developed in our center comprising 12 genes involved in PKD. This panel has been applied to investigate the genetic causes of 32 patients with a clinical suspicion of ADPKD.

RESULTS

We identified a total of 31 variants for 32 individuals, two of which were each detected in two individuals. Twenty-seven out of 31 detected variants were interpreted as pathogenic/likely pathogenic and the remaining 4 of uncertain significance with a molecular diagnostic success rate of 87.5%. Among these variants, 25 PKD1/2 pathogenic/likely pathogenic variants were detected in 32 index patients (78.1%), and variants of uncertain significance in four individuals (12.5% in PKD1/2). The majority of variants was identified in PKD1 (74.2%). Autosomal recessive PKD was identified in one patient, indicating the similarities between recessive and dominant PKD. In concordance with earlier studies, this biallelic PKD1 variant, p.Arg3277Cys, leads to rapidly progressive and severe disease with very early-onset ADPKD.

CONCLUSION

Our findings suggest that targeted gene panel sequencing is expected to be the method of choice to improve diagnostic and prognostic accuracy in PKD patients with heterogeneity in genetic background.