Annotating pathogenic non-coding variants in genic regions

Variant ranking pipeline for complex familial disorders

Identifying genetic susceptibility factors for complex disorders remains a challenging task. To analyze collections of small and large pedigrees where genetic heterogeneity is likely, but biological commonalities are plausible, we have developed a weights-based pipeline to prioritize variants and genes. The Weights-based vAriant Ranking in Pedigrees (WARP) pipeline prioritizes variants using 5 weights: disease incidence rate, number of cases in a family, genome fraction shared amongst cases in a family, allele frequency and variant deleteriousness. Weights, except for the population allele frequency weight, are normalized between 0 and 1. Weights are combined multiplicatively to produce family-specific-variant weights that are then averaged across all families in which the variant is observed to generate a multifamily weight. Sorting multifamily weights in descending order creates a ranked list of variants and genes for further investigation. WARP was validated using familial melanoma sequence data from the European Genome-phenome Archive. The pipeline identified variation in known germline melanoma genes POT1, MITF and BAP1 in 4 out of 13 families (31%). Analysis of the other 9 families identified several interesting genes, some of which might have a role in melanoma. WARP provides an approach to identify disease predisposing genes in studies with small and large pedigrees.

A second hotspot for pathogenic exon-skipping variants in CDC45

Biallelic pathogenic variants in CDC45 are associated with Meier-Gorlin syndrome with craniosynostosis (MGORS type 7), which also includes short stature and absent/hypoplastic patellae. Identified variants act through a hypomorphic loss of function mechanism, to reduce CDC45 activity and impact DNA replication initiation. In addition to missense and premature termination variants, several pathogenic synonymous variants have been identified, most of which cause increased exon skipping of exon 4, which encodes an essential part of the RecJ-orthologue's DHH domain. Here we have identified a second cohort of families segregating CDC45 variants, where patients have craniosynostosis and a reduction in height, alongside common facial dysmorphisms, including thin eyebrows, consistent with MGORS7. Skipping of exon 15 is a consequence of two different variants, including a shared synonymous variant that is enriched in individuals of East Asian ancestry, while other variants in trans are predicted to alter key intramolecular interactions in α/β domain II, or cause retention of an intron within the 3'UTR. Our cohort and functional data confirm exon skipping is a relatively common pathogenic mechanism in CDC45, and highlights the need for alternative splicing events, such as exon skipping, to be especially considered for variants initially predicted to be less likely to cause the phenotype, particularly synonymous variants.

Detecting and understanding meaningful cancerous mutations based on computational models of mRNA splicing

Cancer research has long relied on non-silent mutations. Yet, it has become overwhelmingly clear that silent mutations can affect gene expression and cancer cell fitness. One fundamental mechanism that apparently silent mutations can severely disrupt is alternative splicing. Here we introduce Oncosplice, a tool that scores mutations based on models of proteomes generated using aberrant splicing predictions. Oncosplice leverages a highly accurate neural network that predicts splice sites within arbitrary mRNA sequences, a greedy transcript constructor that considers alternate arrangements of splicing blueprints, and an algorithm that grades the functional divergence between proteins based on evolutionary conservation. By applying this tool to 12M somatic mutations we identify 8K deleterious variants that are significantly depleted within the healthy population; we demonstrate the tool's ability to identify clinically validated pathogenic variants with a positive predictive value of 94%; we show strong enrichment of predicted deleterious mutations across pan-cancer drivers. We also achieve improved patient survival estimation using a proposed set of novel cancer-involved genes. Ultimately, this pipeline enables accelerated insight-gathering of sequence-specific consequences for a class of understudied mutations and provides an efficient way of filtering through massive variant datasets - functionalities with immediate experimental and clinical applications.

Bioinformatics characterization of variants of uncertain significance in pediatric sensorineural hearing loss

Introduction

Rapid advancements in Next Generation Sequencing (NGS) and bioinformatics tools have allowed physicians to obtain genetic testing results in a more rapid, cost-effective, and comprehensive manner than ever before. Around 50% of pediatric sensorineural hearing loss (SNHL) cases are due to a genetic etiology, thus physicians regularly utilize targeted sequencing panels that identify variants in genes related to SNHL. These panels allow for early detection of pathogenic variants which allows physicians to provide anticipatory guidance to families. Molecular testing does not always reveal a clear etiology due to the presence of multigenic variants with varying classifications, including the presence of Variants of Uncertain Significance (VUS). This study aims to perform a preliminary bioinformatics characterization of patients with variants associated with Type II Usher Syndrome in the presence of other multigenic variants. We also provide an interpretation algorithm for physicians reviewing molecular results with medical geneticists.

Methods

Review of records for multigenic and/or VUS results identified several potential subjects of interest. For the purposes of this study, two ADGRV1 compound heterozygotes met inclusion criteria. Sequencing, data processing, and variant calling (the process by which variants are identified from sequence data) was performed at Invitae (San Francisco CA). The preliminary analysis followed the recommendations outlined by the American College of Medical Genetics and Association for Molecular Pathology (ACMG-AMP) in 2015 and 2019. The present study utilizes computational analysis, predictive data, and population data as well as clinical information from chart review and publicly available information in the ClinVar database.

Results

Two subjects were identified as compound heterozygotes for variants in the gene ADGRV1. Subject 1's variants were predicted as deleterious, while Subject 2's variants were predicted as non-deleterious. These results were based on known information of the variants from ClinVar, multiple lines of computational data, population databases, as well as the clinical presentation.

Discussion

Early molecular diagnosis through NGS is ideal, as families are then able to access a wide range of resources that will ultimately support the child as their condition progresses. We recommend that physicians build strong relationships with medical geneticists and carefully review their interpretation before making recommendations to families, particularly when addressing the VUS. Reclassification efforts of VUS are supported by studies like ours that provide evidence of pathogenic or benign effects of variants.

Effect Predictor of Driver Synonymous Mutations Based on Multi-Feature Fusion and Iterative Feature Representation Learning

Accurate identification of driver mutations is crucial in genetic studies of human cancers. While numerous cancer driver missense mutations have been identified, research into potential cancer drivers for synonymous mutations has shown limited success to date. Here, we developed a novel machine learning framework, epSMic, for predicting cancer driver synonymous mutations. epSMic employs an iterative feature representation scheme that facilitates the learning of discriminative features from various sequential models in a supervised iterative mode. We constructed the benchmark datasets and encoded the embedding sequence, physicochemical property, and basic information such as conservation and splicing feature. The evaluation results on benchmark test datasets demonstrate that epSMic outperforms existing methods, making it a valuable tool for researchers in identifying functional synonymous mutations in cancer. We hope epSMic can enable researchers to concentrate on synonymous mutations that have a functional impact on cancer.

Uncovering the clinical relevance of unclassified variants in DNA repair genes: a focus on BRCA negative Tunisian cancer families

Introduction: Recent advances in sequencing technologies have significantly increased our capability to acquire large amounts of genetic data. However, the clinical relevance of the generated data continues to be challenging particularly with the identification of Variants of Uncertain Significance (VUSs) whose pathogenicity remains unclear. In the current report, we aim to evaluate the clinical relevance and the pathogenicity of VUSs in DNA repair genes among Tunisian breast cancer families. Methods: A total of 67 unsolved breast cancer cases have been investigated. The pathogenicity of VUSs identified within 26 DNA repair genes was assessed using different in silico prediction tools including SIFT, PolyPhen2, Align-GVGD and VarSEAK. Effects on the 3D structure were evaluated using the stability predictor DynaMut and molecular dynamics simulation with NAMD. Family segregation analysis was also performed. Results: Among a total of 37 VUSs identified, 11 variants are likely deleterious affecting ATM, BLM, CHEK2, ERCC3, FANCC, FANCG, MSH2, PMS2 and RAD50 genes. The BLM variant, c.3254dupT, is novel and seems to be associated with increased risk of breast, endometrial and colon cancer. Moreover, c.6115G>A in ATM and c.592+3A>T in CHEK2 were of keen interest identified in families with multiple breast cancer cases and their familial cosegregation with disease has been also confirmed. In addition, functional in silico analyses revealed that the ATM variant may lead to protein immobilization and rigidification thus decreasing its activity. We have also shown that FANCC and FANCG variants may lead to protein destabilization and alteration of the structure compactness which may affect FANCC and FANCG protein activity. Conclusion: Our findings revealed that VUSs in DNA repair genes might be associated with increased cancer risk and highlight the need for variant reclassification for better disease management. This will help to improve the genetic diagnosis and therapeutic strategies of cancer patients not only in Tunisia but also in neighboring countries.

Clinical and genetic analyses of a Swedish patient series diagnosed with ataxia

Hereditary ataxia is a heterogeneous group of complex neurological disorders. Next-generation sequencing methods have become a great help in clinical diagnostics, but it may remain challenging to determine if a genetic variant is the cause of the patient's disease. We compiled a consecutive single-center series of 87 patients from 76 families with progressive ataxia of known or unknown etiology. We investigated them clinically and genetically using whole exome or whole genome sequencing. Test methods were selected depending on family history, clinical phenotype, and availability. Genetic results were interpreted based on the American College of Medical Genetics criteria. For high-suspicion variants of uncertain significance, renewed bioinformatical and clinical evaluation was performed to assess the level of pathogenicity. Thirty (39.5%) of the 76 families had received a genetic diagnosis at the end of our study. We present the predominant etiologies of hereditary ataxia in a Swedish patient series. In two families, we established a clinical diagnosis, although the genetic variant was classified as "of uncertain significance" only, and in an additional three families, results are pending. We found a pathogenic variant in one family, but we suspect that it does not explain the complete clinical picture. We conclude that correctly interpreting genetic variants in complex neurogenetic diseases requires genetics and clinical expertise. The neurologist's careful phenotyping remains essential to confirm or reject a diagnosis, also by reassessing clinical findings after a candidate genetic variant is suggested. Collaboration between neurology and clinical genetics and combining clinical and research approaches optimizes diagnostic yield.

Benchmarking splice variant prediction algorithms using massively parallel splicing assays

BACKGROUND

Variants that disrupt mRNA splicing account for a sizable fraction of the pathogenic burden in many genetic disorders, but identifying splice-disruptive variants (SDVs) beyond the essential splice site dinucleotides remains difficult. Computational predictors are often discordant, compounding the challenge of variant interpretation. Because they are primarily validated using clinical variant sets heavily biased to known canonical splice site mutations, it remains unclear how well their performance generalizes.

RESULTS

We benchmark eight widely used splicing effect prediction algorithms, leveraging massively parallel splicing assays (MPSAs) as a source of experimentally determined ground-truth. MPSAs simultaneously assay many variants to nominate candidate SDVs. We compare experimentally measured splicing outcomes with bioinformatic predictions for 3,616 variants in five genes. Algorithms' concordance with MPSA measurements, and with each other, is lower for exonic than intronic variants, underscoring the difficulty of identifying missense or synonymous SDVs. Deep learning-based predictors trained on gene model annotations achieve the best overall performance at distinguishing disruptive and neutral variants, and controlling for overall call rate genome-wide, SpliceAI and Pangolin have superior sensitivity. Finally, our results highlight two practical considerations when scoring variants genome-wide: finding an optimal score cutoff, and the substantial variability introduced by differences in gene model annotation, and we suggest strategies for optimal splice effect prediction in the face of these issues.

CONCLUSION

SpliceAI and Pangolin show the best overall performance among predictors tested, however, improvements in splice effect prediction are still needed especially within exons.

Potential Founder Variants in COL4A4 Identified in Bukharian Jews Linked to Autosomal Dominant and Autosomal Recessive Alport Syndrome

BACKGROUND

Alport syndrome is a hereditary disorder caused by pathogenic variants in the COL4A gene, which can be inherited in an autosomal recessive, dominant, or X-linked pattern. In the Bukharian Jewish population, no founder pathogenic variant has been reported in COL4A4.

METHODS

The cohort included 38 patients from 22 Bukharian Jewish families with suspected Alport syndrome who were referred the nephrogenetics clinic between 2012 and 2022. The study collected demographic, clinical, and genetic data from electronic medical records, which were used to evaluate the molecular basis of the disease using Sanger sequencing, and next-generation sequencing.

RESULTS

Molecular diagnosis was confirmed in 20/38 patients, with each patient having at least one of the three disease-causing COL4A4 variants detected: c.338GA (p.Gly1008Arg), and c.871-6T>C. In addition, two patients were obligate carriers. Overall, there were 17 heterozygotes, 2 compound heterozygotes, and 3 homozygotes. Each variant was detected in more than one unrelated family. All patients had hematuria with/without proteinuria at referral, and the youngest patient with proteinuria (age 5 years) was homozygous for the c.338G>A variant. End-stage renal disease was diagnosed in two patients at the age of 38 years, a compound heterozygote for c.338G>A and c.871-6T>C. Hearing deterioration was detected in three patients, the youngest aged 40 years, all of whom were heterozygous for c.338G>A.

CONCLUSION

This study unveils three novel disease-causing variants, c.3022G>A, c.871-6T>C, and c.338G>A, in the COL4A4 gene that are recurrent among Jews of Bukharian ancestry, and cause Alport syndrome in both dominant and recessive autosomal inheritance patterns.

Defining the single base importance of human mRNAs and lncRNAs

As the fundamental unit of a gene and its transcripts, nucleotides have enormous impacts on the gene function and evolution, and thus on phenotypes and diseases. In order to identify the key nucleotides of one specific gene, it is quite crucial to quantitatively measure the importance of each base on the gene. However, there are still no sequence-based methods of doing that. Here, we proposed Base Importance Calculator (BIC), an algorithm to calculate the importance score of each single base based on sequence information of human mRNAs and long noncoding RNAs (lncRNAs). We then confirmed its power by applying BIC to three different tasks. Firstly, we revealed that BIC can effectively evaluate the pathogenicity of both genes and single bases through single nucleotide variations. Moreover, the BIC score in The Cancer Genome Atlas somatic mutations is able to predict the prognosis of some cancers. Finally, we show that BIC can also precisely predict the transmissibility of SARS-CoV-2. The above results indicate that BIC is a useful tool for evaluating the single base importance of human mRNAs and lncRNAs.

A Combination of BRAF and EZH1/SPOP/ZNF148 Three-Gene Mutational Classifier Improves Benign Call Rate in Indeterminate Thyroid Nodules

Reliable preoperative diagnosis of thyroid nodules remained challenging because of the inconclusiveness of fine-needle aspiration (FNA) cytology. In the present study, 583 formalin-fixed paraffin embedded (FFPE) thyroid nodule tissues and 161 FNA specimens were enrolled retrospectively. Then BRAF V600E, EZH1 Q571R, SPOP P94R, and ZNF148 mutations among these samples were identified using Sanger sequencing. Based on this four-gene genomic classifier, we proposed a two-step modality to diagnose thyroid nodules to differentiate benign and malignant thyroid nodules. In the FFPE group, thyroid cancers were effectively diagnosed in 37.7% (220/583) of neoplasms by the primary BRAF V600E testing, and 15.7% (57/363) of thyroid nodules could be further determined as benign by subsequent EZH1 Q571R, SPOP P94R, and ZNF148 (we called them "ESZ") mutation testing. In the FNA group, 161 BRAF wild-type specimens were classified according to The Bethesda System for Reporting Thyroid Cytopathology (TBSRTC). A total of 7 mutated samples fell within Bethesda categories III-IV, and the mutation rate of "ESZ" in Bethesda III-IV categories was 8.4%. The two-step genomic classifier could further improve thyroid nodule diagnosis, which may inform more optimal patient management.

Predicting molecular mechanisms of hereditary diseases by using their tissue-selective manifestation

How do aberrations in widely expressed genes lead to tissue-selective hereditary diseases? Previous attempts to answer this question were limited to testing a few candidate mechanisms. To answer this question at a larger scale, we developed "Tissue Risk Assessment of Causality by Expression" (TRACE), a machine learning approach to predict genes that underlie tissue-selective diseases and selectivity-related features. TRACE utilized 4,744 biologically interpretable tissue-specific gene features that were inferred from heterogeneous omics datasets. Application of TRACE to 1,031 disease genes uncovered known and novel selectivity-related features, the most common of which was previously overlooked. Next, we created a catalog of tissue-associated risks for 18,927 protein-coding genes (https://netbio.bgu.ac.il/trace/). As proof-of-concept, we prioritized candidate disease genes identified in 48 rare-disease patients. TRACE ranked the verified disease gene among the patient's candidate genes significantly better than gene prioritization methods that rank by gene constraint or tissue expression. Thus, tissue selectivity combined with machine learning enhances genetic and clinical understanding of hereditary diseases.

POU3F3-related disorder: Defining the phenotype and expanding the molecular spectrum

POU3F3 variants cause developmental delay, behavioral problems, hypotonia and dysmorphic features. We investigated the phenotypic and genetic landscape, and genotype-phenotype correlations in individuals with POU3F3-related disorders. We recruited unpublished individuals with POU3F3 variants through international collaborations and obtained updated clinical data on previously published individuals. Trio exome sequencing or single exome sequencing followed by segregation analysis were performed in the novel cohort. Functional effects of missense variants were investigated with 3D protein modeling. We included 28 individuals (5 previously published) from 26 families carrying POU3F3 variants; 23 de novo and one inherited from an affected parent. Median age at study inclusion was 7.4 years. All had developmental delay mainly affecting speech, behavioral difficulties, psychiatric comorbidities and dysmorphisms. Additional features included gastrointestinal comorbidities, hearing loss, ophthalmological anomalies, epilepsy, sleep disturbances and joint hypermobility. Autism, hearing and eye comorbidities, dysmorphisms were more common in individuals with truncating variants, whereas epilepsy was only associated with missense variants. In silico structural modeling predicted that all (likely) pathogenic variants destabilize the DNA-binding region of POU3F3. Our study refined the phenotypic and genetic landscape of POU3F3-related disorders, it reports the functional properties of the identified pathogenic variants, and delineates some genotype-phenotype correlations.

Intronic position +9 and -9 are potentially splicing sites boundary from intronic variants analysis of whole exome sequencing data

Whole exome sequencing (WES) can also detect some intronic variants, which may affect splicing and gene expression, but how to use these intronic variants, and the characteristics about them has not been reported. This study aims to reveal the characteristics of intronic variant in WES data, to further improve the clinical diagnostic value of WES. A total of 269 WES data was analyzed, 688,778 raw variants were called, among these 367,469 intronic variants were in intronic regions flanking exons which was upstream/downstream region of the exon (default is 200 bps). Contrary to expectation, the number of intronic variants with quality control (QC) passed was the lowest at the +2 and -2 positions but not at the +1 and -1 positions. The plausible explanation was that the former had the worst effect on trans-splicing, whereas the latter did not completely abolish splicing. And surprisingly, the number of intronic variants that passed QC was the highest at the +9 and -9 positions, indicating a potential splicing site boundary. The proportion of variants which could not pass QC filtering (false variants) in the intronic regions flanking exons generally accord with "S"-shaped curve. At +5 and -5 positions, the number of variants predicted damaging by software was most. This was also the position at which many pathogenic variants had been reported in recent years. Our study revealed the characteristics of intronic variant in WES data for the first time, we found the +9 and -9 positions might be a potentially splicing sites boundary and +5 and -5 positions were potentially important sites affecting splicing or gene expression, the +2 and -2 positions seem more important splicing site than +1 and -1 positions, and we found variants in intronic regions flanking exons over ± 50 bps may be unreliable. This result can help researchers find more useful variants and demonstrate that WES data is valuable for intronic variants analysis.

Nonsynonymous Synonymous Variants Demand for a Paradigm Shift in Genetics

Synonymous (also known as silent) variations are by definition not considered to change the coded protein. Still many variations in this category affect either protein abundance or properties. As this situation is confusing, we have recently introduced systematics for synonymous variations and those that may on the surface look like synonymous, but these may affect the coded protein in various ways. A new category, unsense variation, was introduced to describe variants that do not introduce a stop codon into the variation site, but which lead to different types of changes in the coded protein. Many of these variations lead to mRNA degradation and missing protein. Here, consequences of the systematics are discussed from the perspectives of variation annotation and interpretation, evolutionary calculations, nonsynonymous-to-synonymous substitution rates, phylogenetics and other evolutionary inferences that are based on the principle of (nearly) neutral synonymous variations. It may be necessary to reassess published results. Further, databases for synonymous variations and prediction methods for such variations should consider unsense variations. Thus, there is a need to evaluate and reflect principles of numerous aspects in genetics, ranging from variation naming and classification to evolutionary calculations.

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.

Automated quantitative trait locus analysis (AutoQTL)

BACKGROUND

Quantitative Trait Locus (QTL) analysis and Genome-Wide Association Studies (GWAS) have the power to identify variants that capture significant levels of phenotypic variance in complex traits. However, effort and time are required to select the best methods and optimize parameters and pre-processing steps. Although machine learning approaches have been shown to greatly assist in optimization and data processing, applying them to QTL analysis and GWAS is challenging due to the complexity of large, heterogenous datasets. Here, we describe proof-of-concept for an automated machine learning approach, AutoQTL, with the ability to automate many complicated decisions related to analysis of complex traits and generate solutions to describe relationships that exist in genetic data.

RESULTS

Using a publicly available dataset of 18 putative QTL from a large-scale GWAS of body mass index in the laboratory rat, Rattus norvegicus, AutoQTL captures the phenotypic variance explained under a standard additive model. AutoQTL also detects evidence of non-additive effects including deviations from additivity and 2-way epistatic interactions in simulated data via multiple optimal solutions. Additionally, feature importance metrics provide different insights into the inheritance models and predictive power of multiple GWAS-derived putative QTL.

CONCLUSIONS

This proof-of-concept illustrates that automated machine learning techniques can complement standard approaches and have the potential to detect both additive and non-additive effects via various optimal solutions and feature importance metrics. In the future, we aim to expand AutoQTL to accommodate omics-level datasets with intelligent feature selection and feature engineering strategies.

PredDSMC: A predictor for driver synonymous mutations in human cancers

Introduction: Driver mutations play a critical role in the occurrence and development of human cancers. Most studies have focused on missense mutations that function as drivers in cancer. However, accumulating experimental evidence indicates that synonymous mutations can also act as driver mutations.

Methods: Here, we proposed a computational method called PredDSMC to accurately predict driver synonymous mutations in human cancers. We first systematically explored four categories of multimodal features, including sequence features, splicing features, conservation scores, and functional scores. Further feature selection was carried out to remove redundant features and improve the model performance. Finally, we utilized the random forest classifier to build PredDSMC.

Results: The results of two independent test sets indicated that PredDSMC outperformed the state-of-the-art methods in differentiating driver synonymous mutations from passenger mutations.

Discussion: In conclusion, we expect that PredDSMC, as a driver synonymous mutation prediction method, will be a valuable method for gaining a deeper understanding of synonymous mutations in human cancers.

Utilization of Targeted RNA-Seq for the Resolution of Variant Pathogenicity and Enhancement of Diagnostic Yield in Dysferlinopathy

For inherited diseases, obtaining a definitive diagnosis is critical for proper disease management, family planning, and participation in clinical trials. This can be challenging for dysferlinopathy due to the significant clinical overlap between the 30+ subtypes of limb–girdle muscular dystrophy (LGMD) and the large number of variants of unknown significance (VUSs) that are identified in the dysferlin gene, DYSF. We performed targeted RNA-Seq using a custom gene-panel in 77 individuals with a clinical/genetic suspicion of dysferlinopathy and evaluated all 111 identified DYSF variants according to the American College of Medical Genetics and Genomics and the Association for Molecular Pathology (ACMG/AMP) guidelines. This evaluation identified 11 novel DYSF variants and allowed for the classification of 87 DYSF variants as pathogenic/likely pathogenic, 8 likely benign, while 16 variants remained VUSs. By the end of the study, 60 of the 77 cases had a definitive diagnosis of dysferlinopathy, which was a 47% increase in diagnostic yield over the rate at study onset. This data shows the ability of RNA-Seq to assist in variant pathogenicity classification and diagnosis of dysferlinopathy and is, therefore, a type of analysis that should be considered when DNA-based genetic analysis is not sufficient to provide a definitive diagnosis.

Deleterious synonymous mutation identification based on selective ensemble strategy

Although previous studies have revealed that synonymous mutations contribute to various human diseases, distinguishing deleterious synonymous mutations from benign ones is still a challenge in medical genomics. Recently, computational tools have been introduced to predict the harmfulness of synonymous mutations. However, most of these computational tools rely on balanced training sets without considering abundant negative samples that could result in deficient performance. In this study, we propose a computational model that uses a selective ensemble to predict deleterious synonymous mutations (seDSM). We construct several candidate base classifiers for the ensemble using balanced training subsets randomly sampled from the imbalanced benchmark training sets. The diversity measures of the base classifiers are calculated by the pairwise diversity metrics, and the classifiers with the highest diversities are selected for integration using soft voting for synonymous mutation prediction. We also design two strategies for filling in missing values in the imbalanced dataset and constructing models using different pairwise diversity metrics. The experimental results show that a selective ensemble based on double fault with the ensemble strategy EKNNI for filling in missing values is the most effective scheme. Finally, using 40-dimensional biology features, we propose a novel model based on a selective ensemble for predicting deleterious synonymous mutations (seDSM). seDSM outperformed other state-of-the-art methods on the independent test sets according to multiple evaluation indicators, indicating that it has an outstanding predictive performance for deleterious synonymous mutations. We hope that seDSM will be useful for studying deleterious synonymous mutations and advancing our understanding of synonymous mutations. The source code of seDSM is freely accessible at https://github.com/xialab-ahu/seDSM.git.

Whole-exome sequencing and bioinformatic analyses revealed differences in gene mutation profiles in papillary thyroid cancer patients with and without benign thyroid goitre background

Background

Papillary thyroid cancer (PTC) is the most common thyroid malignancy. Concurrent presence of cytomorphological benign thyroid goitre (BTG) and PTC lesion is often detected. Aberrant protein profiles were previously reported in patients with and without BTG cytomorphological background. This study aimed to evaluate gene mutation profiles to further understand the molecular mechanism underlying BTG, PTC without BTG background and PTC with BTG background.

Methods

Patients were grouped according to the histopathological examination results: (i) BTG patients (n = 9), (ii) PTC patients without BTG background (PTCa, n = 8), and (iii) PTC patients with BTG background (PTCb, n = 5). Whole-exome sequencing (WES) was performed on genomic DNA extracted from thyroid tissue specimens. Nonsynonymous and splice-site variants with MAF of ≤ 1% in the 1000 Genomes Project were subjected to principal component analysis (PCA). PTC-specific SNVs were filtered against OncoKB and COSMIC while novel SNVs were screened through dbSNP and COSMIC databases. Functional impacts of the SNVs were predicted using PolyPhen-2 and SIFT. Protein-protein interaction (PPI) enrichment of the tumour-related genes was analysed using Metascape and MCODE algorithm.

Results

PCA plots showed distinctive SNV profiles among the three groups. OncoKB and COSMIC database screening identified 36 tumour-related genes including BRCA2 and FANCD2 in all groups. BRAF and 19 additional genes were found only in PTCa and PTCb. "Pathways in cancer", "DNA repair" and "Fanconi anaemia pathway" were among the top networks shared by all groups. However, signalling pathways related to tyrosine kinases were the most significantly enriched in PTCa while "Jak-STAT signalling pathway" and "Notch signalling pathway" were the only significantly enriched in PTCb. Ten SNVs were PTC-specific of which two were novel; DCTN1 c.2786C>G (p.Ala929Gly) and TRRAP c.8735G>C (p.Ser2912Thr). Four out of the ten SNVs were unique to PTCa.

Conclusion

Distinctive gene mutation patterns detected in this study corroborated the previous protein profile findings. We hypothesised that the PTCa and PTCb subtypes differed in the underlying molecular mechanisms involving tyrosine kinase, Jak-STAT and Notch signalling pathways. The potential applications of the SNVs in differentiating the benign from the PTC subtypes requires further validation in a larger sample size.

Hemizygosity can reveal variant pathogenicity on the X-chromosome

Pathogenic variants on the X-chromosome can have more severe consequences for hemizygous males, while heterozygote females can avoid severe consequences due to diploidy and the capacity for nonrandom expression. Thus, when an allele is more common in females this could indicate that it increases the probability of early death in the male hemizygous state, which can be considered a measure of pathogenicity. Importantly, large-scale genomic data now makes it possible to compare allele proportions between the sexes. To discover pathogenic variants on the X-chromosome, we analyzed exome data from 125,748 ancestrally diverse participants in the Genome Aggregation Database (gnomAD). After filtering out duplicates and extremely rare variants, 44,606 of the original 348,221 remained for analysis. We divided the proportion of variant alleles in females by the proportion in males for all variant sites, and then placed each variant into one of three a priori categories: (1) Reference (Primarily synonymous and intronic), (2) Unlikely-to-be-tolerated (Primarily missense), and (3) Least-likely-to-be-tolerated (Primarily frameshift). To assess the impact of ploidy, we compared the distribution of these ratios between pseudoautosomal and non-pseudoautosomal regions. In the non-pseudoautosomal regions, mean female-to-male ratios were lowest among Reference (2.40), greater for Unlikely-to-be-tolerated (2.77) and highest for Least-likely-to-be-tolerated (3.28) variants. Corresponding ratios were lower in the pseudoautosomal regions (1.52, 1.57, and 1.68, respectively), with the most extreme ratio being just below 11. Because pathogenic effects in the pseudoautosomal regions should not drive ratio increases, this maximum ratio provides an upper bound for baseline noise. In the non-pseudoautosomal regions, 319 variants had a ratio over 11. In sum, we identified a measure with a dataset specific threshold for identifying pathogenicity in non-pseudoautosomal X-chromosome variants: the female-to-male allele proportion ratio.

frDSM: An Ensemble Predictor With Effective Feature Representation for Deleterious Synonymous Mutation in Human Genome

With the discovery of causality between synonymous mutations and diseases, it has become increasingly important to identify deleterious synonymous mutations for better understanding of their functional mechanisms. Although several machine learning methods have been proposed to solve the task, an effective feature representation method that can make use of the inner difference and relevance between deleterious and benign synonymous mutations is still challenging considering the vast number of synonymous mutations in human genome. In this work, we developed a robust and accurate predictor called frDSM for deleterious synonymous mutation prediction using logistic regression. More specifically, we introduced an effective feature representation learning method which exploits multiple feature descriptors from different perspectives including functional scores obtained from previously computational methods, evolutionary conservation, splicing and sequence feature descriptors, and these features descriptors were input into the 76 XGBoost classifiers to obtain the predictive probabilities values. These probabilities were concatenated to generate the 76-dimension new feature vector, and feature selection method was used to remove redundant and irrelevant features. Experimental results show that frDSM enables robust and accurate prediction than the competing prediction methods with 31 optimal features, which demonstrated the effectiveness of the feature representation learning method. frDSM is freely available at http://frdsm.xialab.info.

Computational prediction of human deep intronic variation

BACKGROUND

The adoption of whole-genome sequencing in genetic screens has facilitated the detection of genetic variation in the intronic regions of genes, far from annotated splice sites. However, selecting an appropriate computational tool to discriminate functionally relevant genetic variants from those with no effect is challenging, particularly for deep intronic regions where independent benchmarks are scarce.

RESULTS

In this study, we have provided an overview of the computational methods available and the extent to which they can be used to analyze deep intronic variation. We leveraged diverse datasets to extensively evaluate tool performance across different intronic regions, distinguishing between variants that are expected to disrupt splicing through different molecular mechanisms. Notably, we compared the performance of SpliceAI, a widely used sequence-based deep learning model, with that of more recent methods that extend its original implementation. We observed considerable differences in tool performance depending on the region considered, with variants generating cryptic splice sites being better predicted than those that potentially affect splicing regulatory elements. Finally, we devised a novel quantitative assessment of tool interpretability and found that tools providing mechanistic explanations of their predictions are often correct with respect to the ground - information, but the use of these tools results in decreased predictive power when compared to black box methods.

CONCLUSIONS

Our findings translate into practical recommendations for tool usage and provide a reference framework for applying prediction tools in deep intronic regions, enabling more informed decision-making by practitioners.

Mutational analysis of phospholipase C epsilon 1 gene in Egyptian children with steroid-resistant nephrotic syndrome

Background

Steroid-resistant nephrotic syndrome (SRNS) is characterized by unresponsiveness of nephrotic range proteinuria to standard steroid therapy, and is the main cause of childhood renal failure. The identification of more than 53 monogenic causes of SRNS has led researchers to focus on the genetic mutations related to the molecular mechanisms of the disease. Mutations in the PLCE1 gene, which encodes phospholipase C epsilon 1 (PLCε1), have been described in patients with early-onset SRNS characterized by progressive renal failure. In this study we screened for PLCE1 mutations in Egyptian children with SRNS. This is a descriptive case series study aiming to screen for PLCE1 gene mutations by direct sequencing of five exons—9, 12, 15, 19, 27—in 20 Egyptian children with SRNS who entered the Nephrology Unit, Faculty of Medicine, Ain-Shams University from November 2015 to December 2017. The variants detected were submitted to in silico analysis.

Results

We screened for mutations in five selected exons of PLCE1 gene. We identified seven variants in the five selected exons with homozygous and heterozygous inheritance pattern, two are intronic variants, two are silent variants, and three are missense variants. We identified four novel variants two are silent with no clinical significance and two are missense with uncertain clinical significance and pathogenic in-silico predictions; one p.Arg1230His in exon 12, the other is p.Glu1393Lys in exon 15.

Conclusions

We identified four novel mutations, findings which added to the registered SNP spectrum of the PLCE1 gene. These results widen the spectrum of PLCE1 gene mutations and support the importance of genetic testing in different populations of SRNS patients, therefore, to assess the vulnerability of Egyptian children to SRNS candidate genes, further studies needed on a larger number of cases which undoubtedly provide new insights into the pathogenic mechanisms of SRNS and might help in control of the patient. Additionally, the use of computational scoring and modeling tools may assist in the evaluation of the way in which the SNPs affect protein functionality.

Transcriptome analysis provides critical answers to the "variants of uncertain significance" conundrum

While whole-genome and exome sequencing have transformed our collective understanding of genetics' role in disease pathogenesis, there are certain conditions and populations for whom DNA-level data fails to identify the underlying genetic etiology. Specifically, patients of non-White race and non-European ancestry are disproportionately affected by "variants of unknown/uncertain significance" (VUS), limiting the scope of precision medicine for minority patients and perpetuating health disparities. VUS often include deep intronic and splicing variants which are difficult to interpret from DNA data alone. RNA analysis can illuminate the consequences of VUS, thereby allowing for their reclassification as pathogenic versus benign. Here we review the critical role transcriptome analysis plays in clarifying VUS in both neoplastic and non-neoplastic diseases.

Polymorphisms of nicotinic acetylcholine receptors in Alzheimer’s disease: a systematic review and data analysis

Objective

This study was conducted to accomplish a better insight into the impact of single nucleotide polymorphisms (SNPs) of nicotinic acetylcholine receptors (nAChR) at the risk of Alzheimer’s disease (AD) and their possible pathogenicity.

Methods

We carried out a systemic review of accessible studies. The case–control studies were assessed by an electronic search of international and local databases to identify relevant studies on SNPs relating to nAChR genes in AD. Two reviewers evaluated the inclusion/exclusion criteria, summarized, and analyzed the extracted data. We used odds ratios (ORs) with 95% confidence intervals (CIs) for reporting our data. Online databases were checked for possible pathogenicity of statistically significant SNPs. Also, online databases, including NCBI, NIH, ClinVar, RegulomeDB, and Ensemble, were used to analyze and identify structure and function, DNA features, and flank sequencing in SNPs.

Results

Among all collected SNPs, rs4779978 and rs1827294 on CHRNA7, rs1044394 on CHRNA4, and rs1127314 on CHRNB2 showed statistically significant between AD cases and controls.

Conclusions

Some SNPs from the reviewed reports show evidence supporting their possible involvement in AD pathology. However, more comprehensive studies are necessary to identify the exact correlation and their role on the pathogenicity of disease.

When a Synonymous Variant Is Nonsynonymous

Term synonymous variation is widely used, but frequently in a wrong or misleading meaning and context. Twenty three point eight % of possible nucleotide substitution types in the universal genetic code are for synonymous amino acid changes, but when these variants have a phenotype and functional effect, they are very seldom synonymous. Such variants may manifest changes at DNA, RNA and/or protein levels. Large numbers of variations are erroneously annotated as synonymous, which causes problems e.g., in clinical genetics and diagnosis of diseases. To facilitate precise communication, novel systematics and nomenclature are introduced for variants that when looking only at the genetic code seem like synonymous, but which have phenotypes. A new term, unsense variant is defined as a substitution in the mRNA coding region that affects gene expression and protein production without introducing a stop codon in the variation site. Such variants are common and need to be correctly annotated. Proper naming and annotation are important also to increase awareness of these variants and their consequences.

Performance evaluation of computational methods for splice-disrupting variants and improving the performance using the machine learning-based framework

A critical challenge in genetic diagnostics is the assessment of genetic variants associated with diseases, specifically variants that fall out with canonical splice sites, by altering alternative splicing. Several computational methods have been developed to prioritize variants effect on splicing; however, performance evaluation of these methods is hampered by the lack of large-scale benchmark datasets. In this study, we employed a splicing-region-specific strategy to evaluate the performance of prediction methods based on eight independent datasets. Under most conditions, we found that dbscSNV-ADA performed better in the exonic region, S-CAP performed better in the core donor and acceptor regions, S-CAP and SpliceAI performed better in the extended acceptor region and MMSplice performed better in identifying variants that caused exon skipping. However, it should be noted that the performances of prediction methods varied widely under different datasets and splicing regions, and none of these methods showed the best overall performance with all datasets. To address this, we developed a new method, machine learning-based classification of splice sites variants (MLCsplice), to predict variants effect on splicing based on individual methods. We demonstrated that MLCsplice achieved stable and superior prediction performance compared with any individual method. To facilitate the identification of the splicing effect of variants, we provided precomputed MLCsplice scores for all possible splice sites variants across human protein-coding genes (http://39.105.51.3:8090/MLCsplice/). We believe that the performance of different individual methods under eight benchmark datasets will provide tentative guidance for appropriate method selection to prioritize candidate splice-disrupting variants, thereby increasing the genetic diagnostic yield.

Polygenic risk impacts PDGFRA mutation penetrance in non-syndromic cleft lip and palate

Non-syndromic cleft lip with or without cleft palate (NSCL/P) is a common, severe craniofacial malformation that imposes significant medical, psychosocial and financial burdens. NSCL/P is a multifactorial disorder with genetic and environmental factors playing etiologic roles. Currently, only 25% of the genetic variation underlying NSCL/P has been identified by linkage, candidate gene and genome-wide association studies. In this study, whole-genome sequencing and genome-wide genotyping followed by polygenic risk score (PRS) and linkage analyses were used to identify the genetic etiology of NSCL/P in a large three-generation family. We identified a rare missense variant in PDGFRA (c.C2740T; p.R914W) as potentially etiologic in a gene-based association test using pVAAST (P = 1.78 × 10-4) and showed decreased penetrance. PRS analysis suggested that variant penetrance was likely modified by common NSCL/P risk variants, with lower scores found among unaffected carriers. Linkage analysis provided additional support for PRS-modified penetrance, with a 7.4-fold increase in likelihood after conditioning on PRS. Functional characterization experiments showed that the putatively causal variant was null for signaling activity in vitro; further, perturbation of pdgfra in zebrafish embryos resulted in unilateral orofacial clefting. Our findings show that a rare PDGFRA variant, modified by additional common NSCL/P risk variants, have a profound effect on NSCL/P risk. These data provide compelling evidence for multifactorial inheritance long postulated to underlie NSCL/P and may explain some unusual familial patterns.

Establishment and Validation of a Ferroptosis-Related lncRNA Signature for Prognosis Prediction in Lower-Grade Glioma

Background

The prognosis of lower-grade glioma (LGG) is highly variable, and more accurate predictors are still needed. The aim of our study was to explore the prognostic value of ferroptosis-related long non-coding RNAs (lncRNAs) in LGG and to develop a novel risk signature for predicting survival with LGG.

Methods

We first integrated multiple datasets to screen for prognostic ferroptosis-related lncRNAs in LGG. A least absolute shrinkage and selection operator (LASSO) analysis was then utilized to develop a risk signature for prognostic prediction. Based on the results of multivariate Cox analysis, a prognostic nomogram model for LGG was constructed. Finally, functional enrichment analysis, single-sample gene set enrichment analysis (ssGSEA), immunity, and m6A correlation analyses were conducted to explore the possible mechanisms by which these ferroptosis-related lncRNAs affect survival with LGG.

Results

A total of 11 ferroptosis-related lncRNAs related to the prognosis of LGG were identified. Based on prognostic lncRNAs, a risk signature consisting of 8 lncRNAs was constructed and demonstrated good predictive performance in both the training and validation cohorts. Correlation analysis suggested that the risk signature was closely linked to clinical features. The nomogram model we constructed by combining the risk signature and clinical parameters proved to be more accurate in predicting the prognosis of LGG. In addition, there were differences in the levels of immune cell infiltration, immune-related functions, immune checkpoints, and m6A-related gene expression between the high- and low-risk groups.

Conclusion

In summary, our ferroptosis-related lncRNA signature exhibits good performance in predicting the prognosis of LGG. This study may provide useful insight into the treatment of LGG.

Targeted RNAseq Improves Clinical Diagnosis of Very Early-Onset Pediatric Immune Dysregulation

Despite increased use of whole exome sequencing (WES) for the clinical analysis of rare disease, overall diagnostic yield for most disorders hovers around 30%. Previous studies of mRNA have succeeded in increasing diagnoses for clearly defined disorders of monogenic inheritance. We asked if targeted RNA sequencing could provide similar benefits for primary immunodeficiencies (PIDs) and very early-onset inflammatory bowel disease (VEOIBD), both of which are difficult to diagnose due to high heterogeneity and variable severity. We performed targeted RNA sequencing of a panel of 260 immune-related genes for a cohort of 13 patients (seven suspected PID cases and six VEOIBD) and analyzed variants, splicing, and exon usage. Exonic variants were identified in seven cases, some of which had been previously prioritized by exome sequencing. For four cases, allele specific expression or lack thereof provided additional insights into possible disease mechanisms. In addition, we identified five instances of aberrant splicing associated with four variants. Three of these variants had been previously classified as benign in ClinVar based on population frequency. Digenic or oligogenic inheritance is suggested for at least two patients. In addition to validating the use of targeted RNA sequencing, our results show that rare disease research will benefit from incorporating contributing genetic factors into the diagnostic approach.

Case Report: MYO5B Homozygous Variant c.2090+3A>T Causes Intron Retention Related to Chronic Cholestasis and Diarrhea

Background: Biallelically mutated MYO5B is associated with microvillus inclusion disease (MVID, MIM: 251850), cholestasis, or both. This study aims at validating the splicing alteration and clinical features of an intron variant for diagnosis.

Case Presentation: A homozygous variant of MYO5B, NM_001080467.2:c.2090+3A > T (NP_001073936.1:p.?) in intron 17, was identified in a patient suffering from chronic cholestasis and diarrhea. Functional validation showed that this variant caused 185 bp of intron retention in its mRNA and was predicted to present a premature translation termination site for myoVb (p.Arg697fs*47) in the head motor domain. In addition, bowel biopsy revealed decreased microvilli and local lesions of microvillus inclusion in the duodena of the patient. The patient was presented with neonatal cholestasis leading to cirrhosis, intractable diarrhea, cholelithiasis, hepatic cyst, corneal opacity, and failure to thrive.

Conclusion: Our study demonstrated an intronic homozygous variant of MYO5B that affected an intron, subsequently altering splicing and leading to combined cholestasis and MVID. Our results further supported the underlying genotype–phenotype correlations and extended clinical practices toward its diagnosis and management.

Genetic Analysis of Coronary Artery Disease Using Tree-Based Automated Machine Learning Informed By Biology-Based Feature Selection

Machine Learning (ML) approaches are increasingly being used in biomedical applications. Important challenges of ML include choosing the right algorithm and tuning the parameters for optimal performance. Automated ML (AutoML) methods, such as Tree-based Pipeline Optimization Tool (TPOT), have been developed to take some of the guesswork out of ML thus making this technology available to users from more diverse backgrounds. The goals of this study were to assess applicability of TPOT to genomics and to identify combinations of single nucleotide polymorphisms (SNPs) associated with coronary artery disease (CAD), with a focus on genes with high likelihood of being good CAD drug targets. We leveraged public functional genomic resources to group SNPs into biologically meaningful sets to be selected by TPOT. We applied this strategy to data from the U.K. Biobank, detecting a strikingly recurrent signal stemming from a group of 28 SNPs. Importance analysis of these SNPs uncovered functional relevance of the top SNPs to genes whose association with CAD is supported in the literature and other resources. Furthermore, we employed game-theory based metrics to study SNP contributions to individual-level TPOT predictions and discover distinct clusters of well-predicted CAD cases. The latter indicates a promising approach towards precision medicine.

Genome interpretation using in silico predictors of variant impact

Estimating the effects of variants found in disease driver genes opens the door to personalized therapeutic opportunities. Clinical associations and laboratory experiments can only characterize a tiny fraction of all the available variants, leaving the majority as variants of unknown significance (VUS). In silico methods bridge this gap by providing instant estimates on a large scale, most often based on the numerous genetic differences between species. Despite concerns that these methods may lack reliability in individual subjects, their numerous practical applications over cohorts suggest they are already helpful and have a role to play in genome interpretation when used at the proper scale and context. In this review, we aim to gain insights into the training and validation of these variant effect predicting methods and illustrate representative types of experimental and clinical applications. Objective performance assessments using various datasets that are not yet published indicate the strengths and limitations of each method. These show that cautious use of in silico variant impact predictors is essential for addressing genome interpretation challenges.

Synonymous Variants: Necessary Nuance in our Understanding of Cancer Drivers and Treatment Outcomes

Once called "silent mutations" and assumed to have no effect on protein structure and function, synonymous variants are now recognized to be drivers for some cancers. There have been significant advances in our understanding of the numerous mechanisms by which synonymous single nucleotide variants (sSNVs) can affect protein structure and function by affecting pre-mRNA splicing, mRNA expression, stability, folding, miRNA binding, translation kinetics, and co-translational folding. This review highlights the need for considering sSNVs in cancer biology to gain a better understanding of the genetic determinants of human cancers and to improve their diagnosis and treatment. We surveyed the literature for reports of sSNVs in cancer and found numerous studies on the consequences of sSNVs on gene function with supporting in vitro evidence. We also found reports of sSNVs that have statistically significant associations with specific cancer types but for which in vitro studies are lacking to support the reported associations. Additionally, we found reports of germline and somatic sSNVs that were observed in numerous clinical studies and for which in silico analysis predicts possible effects on gene function. We provide a review of these investigations and discuss necessary future studies to elucidate the mechanisms by which sSNVs disrupt protein function and are play a role in tumorigeneses, cancer progression, and treatment efficacy. As splicing dysregulation is one of the most well recognized mechanisms by which sSNVs impact protein function, we also include our own in silico analysis for predicting which sSNVs may disrupt pre-mRNA splicing.

National Genome Initiatives in Europe and the United Kingdom in the Era of Whole-Genome Sequencing: A Comprehensive Review

Identification of genomic variability in population plays an important role in the clinical diagnostics of human genetic diseases. Thanks to rapid technological development in the field of massive parallel sequencing technologies, also known as next-generation sequencing (NGS), complex genomic analyses are now easier and cheaper than ever before, which consequently leads to more effective utilization of these techniques in clinical practice. However, interpretation of data from NGS is still challenging due to several issues caused by natural variability of DNA sequences in human populations. Therefore, development and realization of projects focused on description of genetic variability of local population (often called "national or digital genome") with a NGS technique is one of the best approaches to address this problem. The next step of the process is to share such data via publicly available databases. Such databases are important for the interpretation of variants with unknown significance or (likely) pathogenic variants in rare diseases or cancer or generally for identification of pathological variants in a patient's genome. In this paper, we have compiled an overview of published results of local genome sequencing projects from United Kingdom and Europe together with future plans and perspectives for newly announced ones.

The benefit of diagnostic whole genome sequencing in schizophrenia and other psychotic disorders

Schizophrenia has a multifactorial etiology, involving a polygenic architecture. The potential benefit of whole genome sequencing (WGS) in schizophrenia and other psychotic disorders is not well studied. We investigated the yield of clinical WGS analysis in 251 families with a proband diagnosed with schizophrenia (N = 190), schizoaffective disorder (N = 49), or other conditions involving psychosis (N = 48). Participants were recruited in Israel and USA, mainly of Jewish, Arab, and other European ancestries. Trio (parents and proband) WGS was performed for 228 families (90.8%); in the other families, WGS included parents and at least two affected siblings. In the secondary analyses, we evaluated the contribution of rare variant enrichment in particular gene sets, and calculated polygenic risk score (PRS) for schizophrenia. For the primary outcome, diagnostic rate was 6.4%; we found clinically significant, single nucleotide variants (SNVs) or small insertions or deletions (indels) in 14 probands (5.6%), and copy number variants (CNVs) in 2 (0.8%). Significant enrichment of rare loss-of-function variants was observed in a gene set of top schizophrenia candidate genes in affected individuals, compared with population controls (N = 6,840). The PRS for schizophrenia was significantly increased in the affected individuals group, compared to their unaffected relatives. Last, we were also able to provide pharmacogenomics information based on CYP2D6 genotype data for most participants, and determine their antipsychotic metabolizer status. In conclusion, our findings suggest that WGS may have a role in the setting of both research and genetic counseling for individuals with schizophrenia and other psychotic disorders and their families.

PIGN encephalopathy: Characterizing the epileptology

OBJECTIVE

Epilepsy is common in patients with PIGN diseases due to biallelic variants; however, limited epilepsy phenotyping data have been reported. We describe the epileptology of PIGN encephalopathy.

METHODS

We recruited patients with epilepsy due to biallelic PIGN variants and obtained clinical data regarding age at seizure onset/offset and semiology, development, medical history, examination, electroencephalogram, neuroimaging, and treatment. Seizure and epilepsy types were classified.

RESULTS

Twenty six patients (13 female) from 26 families were identified, with mean age 7 years (range = 1 month to 21 years; three deceased). Abnormal development at seizure onset was present in 25 of 26. Developmental outcome was most frequently profound (14/26) or severe (11/26). Patients presented with focal motor (12/26), unknown onset motor (5/26), focal impaired awareness (1/26), absence (2/26), myoclonic (2/26), myoclonic-atonic (1/26), and generalized tonic-clonic (2/26) seizures. Twenty of 26 were classified as developmental and epileptic encephalopathy (DEE): 55% (11/20) focal DEE, 30% (6/20) generalized DEE, and 15% (3/20) combined DEE. Six had intellectual disability and epilepsy (ID+E): two generalized and four focal epilepsy. Mean age at seizure onset was 13 months (birth to 10 years), with a lower mean onset in DEE (7 months) compared with ID+E (33 months). Patients with DEE had drug-resistant epilepsy, compared to 4/6 ID+E patients, who were seizure-free. Hyperkinetic movement disorder occurred in 13 of 26 patients. Twenty-seven of 34 variants were novel. Variants were truncating (n = 7), intronic and predicted to affect splicing (n = 7), and missense or inframe indels (n = 20, of which 11 were predicted to affect splicing). Seven variants were recurrent, including p.Leu311Trp in 10 unrelated patients, nine with generalized seizures, accounting for nine of the 11 patients in this cohort with generalized seizures.

SIGNIFICANCE

PIGN encephalopathy is a complex autosomal recessive disorder associated with a wide spectrum of epilepsy phenotypes, typically with substantial profound to severe developmental impairment.

Whole exome sequencing in Alopecia Areata identifies rare variants in KRT82

Alopecia areata is a complex genetic disease that results in hair loss due to the autoimmune-mediated attack of the hair follicle. We previously defined a role for both rare and common variants in our earlier GWAS and linkage studies. Here, we identify rare variants contributing to Alopecia Areata using a whole exome sequencing and gene-level burden analyses approach on 849 Alopecia Areata patients compared to 15,640 controls. KRT82 is identified as an Alopecia Areata risk gene with rare damaging variants in 51 heterozygous Alopecia Areata individuals (6.01%), achieving genome-wide significance (p = 2.18E−07). KRT82 encodes a hair-specific type II keratin that is exclusively expressed in the hair shaft cuticle during anagen phase, and its expression is decreased in Alopecia Areata patient skin and hair follicles. Finally, we find that cases with an identified damaging KRT82 variant and reduced KRT82 expression have elevated perifollicular CD8 infiltrates. In this work, we utilize whole exome sequencing to successfully identify a significant Alopecia Areata disease-relevant gene, KRT82, and reveal a proposed mechanism for rare variant predisposition leading to disrupted hair shaft integrity.

Common variants have been discovered to be associated with Alopecia Areata; however, rare variants have been less well studied. Here, the authors use whole-exome sequencing to identify associated rare variants in the hair keratin gene KRT82. Further, they find that individuals with Alopecia Areata have reduced expression of KRT82 in the skin and hair follicle.

Case Report: Biallelic Loss of Function ATM due to Pathogenic Synonymous and Novel Deep Intronic Variant c.1803-270T > G Identified by Genome Sequencing in a Child With Ataxia–Telangiectasia

Ataxia–telangiectasia (AT) is a complex neurodegenerative disease with an increased risk for bone marrow failure and malignancy. AT is caused by biallelic loss of function variants in ATM, which encodes a phosphatidylinositol 3-kinase that responds to DNA damage. Herein, we report a child with progressive ataxia, chorea, and genome instability, highly suggestive of AT. The clinical ataxia gene panel identified a maternal heterozygous synonymous variant (NM_000051.3: c.2250G > A), previously described to result in exon 14 skipping. Subsequently, trio genome sequencing led to the identification of a novel deep intronic variant [NG_009830.1(NM_000051.3): c.1803-270T > G] inherited from the father. Transcript analyses revealed that c.1803-270T > G results in aberrant inclusion of 56 base pairs of intron 11. In silico tests predicted a premature stop codon as a consequence, suggesting non-functional ATM; and DNA repair analyses confirmed functional loss of ATM. Our findings highlight the power of genome sequencing, considering deep intronic variants in undiagnosed rare disease patients.

KDM6A missense variants hamper H3 histone demethylation in lung squamous cell carcinoma

KDM6A is the disease causative gene of type 2 Kabuki Syndrome, a rare multisystem disease; it is also a known cancer driver gene, with multiple somatic mutations found in a few cancer types. In this study, we looked at eleven missense variants in lung squamous cell carcinoma, one of the most common lung cancer subtypes, to see how they affect the KDM6A catalytic mechanisms. We found that they influence the interaction with histone H3 and the exposure of the trimethylated Lys27, which is critical for wild-type physiological function to varying degrees, by altering the conformational transition.

Decoding the effects of synonymous variants

Synonymous single nucleotide variants (sSNVs) are common in the human genome but are often overlooked. However, sSNVs can have significant biological impact and may lead to disease. Existing computational methods for evaluating the effect of sSNVs suffer from the lack of gold-standard training/evaluation data and exhibit over-reliance on sequence conservation signals. We developed synVep (synonymous Variant effect predictor), a machine learning-based method that overcomes both of these limitations. Our training data was a combination of variants reported by gnomAD (observed) and those unreported, but possible in the human genome (generated). We used positive-unlabeled learning to purify the generated variant set of any likely unobservable variants. We then trained two sequential extreme gradient boosting models to identify subsets of the remaining variants putatively enriched and depleted in effect. Our method attained 90% precision/recall on a previously unseen set of variants. Furthermore, although synVep does not explicitly use conservation, its scores correlated with evolutionary distances between orthologs in cross-species variation analysis. synVep was also able to differentiate pathogenic vs. benign variants, as well as splice-site disrupting variants (SDV) vs. non-SDVs. Thus, synVep provides an important improvement in annotation of sSNVs, allowing users to focus on variants that most likely harbor effects.

Genomic Variation Prediction: A Summary From Different Views

Structural variations in the genome are closely related to human health and the occurrence and development of various diseases. To understand the mechanisms of diseases, find pathogenic targets, and carry out personalized precision medicine, it is critical to detect such variations. The rapid development of high-throughput sequencing technologies has accelerated the accumulation of large amounts of genomic mutation data, including synonymous mutations. Identifying pathogenic synonymous mutations that play important roles in the occurrence and development of diseases from all the available mutation data is of great importance. In this paper, machine learning theories and methods are reviewed, efficient and accurate pathogenic synonymous mutation prediction methods are developed, and a standardized three-level variant analysis framework is constructed. In addition, multiple variation tolerance prediction models are studied and integrated, and new ideas for structural variation detection based on deep information mining are explored.

CD8+ cell somatic mutations in multiple sclerosis patients and controls-Enrichment of mutations in STAT3 and other genes implicated in hematological malignancies

Somatic mutations have a central role in cancer but their role in other diseases such as common autoimmune disorders is not clear. Previously we and others have demonstrated that especially CD8+ T cells in blood can harbor persistent somatic mutations in some patients with multiple sclerosis (MS) and rheumatoid arthritis. Here we concentrated on CD8+ cells in more detail and tested (i) how commonly somatic mutations are detectable, (ii) does the overall mutation load differ between MS patients and controls, and (iii) do the somatic mutations accumulate non-randomly in certain genes? We separated peripheral blood CD8+ cells from newly diagnosed relapsing MS patients (n = 21) as well as matched controls (n = 21) and performed next-generation sequencing of the CD8+ cells' DNA, limiting our search to a custom panel of 2524 immunity and cancer related genes, which enabled us to obtain a median sequencing depth of over 2000x. We discovered nonsynonymous somatic mutations in all MS patients' and controls' CD8+ cell DNA samples, with no significant difference in number between the groups (p = 0.60), at a median allelic fraction of 0.5% (range 0.2-8.6%). The mutations showed statistically significant clustering especially to the STAT3 gene, and also enrichment to the SMARCA2, DNMT3A, SOCS1 and PPP3CA genes. Known activating STAT3 mutations were found both in MS patients and controls and overall 1/5 of the mutations were previously described cancer mutations. The detected clustering suggests a selection advantage of the mutated CD8+ clones and calls for further research on possible phenotypic effects.

Refinement of pathogenicity classification of variants associated with familial hypercholesterolemia: Implications for clinical diagnosis

BACKGROUND

The lack of functional evidence for most variants detected during the molecular screening of patients with clinical familial hypercholesterolemia (FH) makes the definitive diagnosis difficult.

METHODS

A total of 552 variants in LDLR, APOB, PCSK9 and LDLRAP1 genes found in 449 mutation-positive FH (FH/M+) patients were considered. Pathogenicity update was performed following the American College of Medical Genetics and Genomics (ACMG) guidelines with additional specifications on copy number variants, functional studies, in silico prediction and co-segregation criteria for LDLR, APOB and PCSK9 genes. Pathogenicity of LDLRAP1 variants was updated by using ACMG criteria with no change to original scoring.

RESULTS

After reclassification, the proportion of FH/M+ carriers of pathogenic (P) or likely pathogenic (LP) variants, and FH/M+ carriers of likely benign (LB) or benign (B) variants, was higher than that defined by standard criteria (81.5% vs. 79.7% and 7.1% vs. 2.7%). The refinement of pathogenicity classification also reduced the percentage of FH with variants of uncertain significance (VUS) (17.7% vs. 11.4%). After adjustment, the FH diagnosis by refined criteria best predicted LDL-C levels (P_adj <0.001). Notably, FH with VUS variants had higher LDL-C than those with LB (all P_adj ≤ 0.033), but similar to those with LP variants.

CONCLUSION

Accurate variant interpretation best predicts the increase of LDL-C levels and shows its clinical utility in the molecular diagnosis of FH.

Accurate interpretation of genetic variants in sudden unexpected death in infancy by trio-targeted gene-sequencing panel analysis

In sudden unexpected death in infancy cases, postmortem genetic analysis with next-generation sequencing potentially can extract candidate genes associated with sudden death. However, it is difficult to accurately interpret the clinically significant genetic variants. The study aim was to conduct trio analysis of cases of sudden unexpected death in infancy and their parents to more accurately interpret the clinically significant disease-associated gene variants associated with cause of death. From the TruSight One panel targeting 4813 genes we extracted candidate genetic variants of 66 arrhythmia-, 63 inherited metabolic disease-, 81 mitochondrial disease-, and 6 salt-losing tubulopathy-related genes in 7 cases and determined if they were de novo or parental-derived variants. Thirty-four parental-derived variants and no de novo variants were found, but none appeared to be related to the cause of death. Using trio analysis and an in silico algorithm to analyze all 4813 genes, we identified OBSCN of compound heterozygous and HCCS of hemizygous variants as new candidate genetic variants related to cause of death. Genetic analysis of these deceased infants and their living parents can provide more accurate interpretation of the clinically significant genetic variants than previously possible and help confirm the cause of death.

The promise of automated machine learning for the genetic analysis of complex traits

The genetic analysis of complex traits has been dominated by parametric statistical methods due to their theoretical properties, ease of use, computational efficiency, and intuitive interpretation. However, there are likely to be patterns arising from complex genetic architectures which are more easily detected and modeled using machine learning methods. Unfortunately, selecting the right machine learning algorithm and tuning its hyperparameters can be daunting for experts and non-experts alike. The goal of automated machine learning (AutoML) is to let a computer algorithm identify the right algorithms and hyperparameters thus taking the guesswork out of the optimization process. We review the promises and challenges of AutoML for the genetic analysis of complex traits and give an overview of several approaches and some example applications to omics data. It is our hope that this review will motivate studies to develop and evaluate novel AutoML methods and software in the genetics and genomics space. The promise of AutoML is to enable anyone, regardless of training or expertise, to apply machine learning as part of their genetic analysis strategy.

ASTL is mutated in female infertility

Female infertility is a relatively common phenotype with a growing number of single gene causes although these account for only a minority of cases. Here, we report a consanguineous family in which adult females who are homozygous for a truncating variant in ASTL display markedly reduced fertility in a pattern strikingly similar to Astl^-/- female mice. ASTL encodes ovastacin, which is known to trigger zona pellucida hardening (ZPH) as part of the cortical reaction upon fertilization. ZPH is required for normal early embryonic development and its absence can be caused by pathogenic variants in other zona pellucida proteins that result in a similar infertility phenotype in humans and mouse. This is the first report of ASTL-related infertility in humans and suggests that the inclusion of ASTL in female infertility gene panels is warranted.