Gene-specific function prediction for non-synonymous mutations in monogenic diabetes genes

Novel gene-specific Bayesian Gaussian mixture model to predict the missense variants pathogenicity of Sanfilippo syndrome

MPS III is an autosomal recessive lysosomal storage disease caused mainly by missense variants in the NAGLU, GNS, HGSNAT, and SGSH genes. The pathogenicity interpretation of missense variants is still challenging. We aimed to develop unsupervised clustering-based pathogenicity predictor scores using extracted features from eight in silico predictors to predict the impact of novel missense variants of Sanfilippo syndrome. The model was trained on a dataset consisting of 415 uncertain significant (VUS) missense NAGLU variants. Performance The SanfilippoPred tool was evaluated by validation and test datasets consisting of 197-labelled NAGLU missense variants, and its performance was compared versus individual pathogenicity predictors using receiver operating characteristic (ROC) analysis. Moreover, we tested the SanfilippoPred tool using extra-labelled 427 missense variants to assess its specificity and sensitivity threshold. Application of the trained machine learning (ML) model on the test dataset of labelled NAGLU missense variants showed that SanfilippoPred has an accuracy of 0.93 (0.86-0.97 at CI 95%), sensitivity of 0.93, and specificity of 0.92. The comparative performance of the SanfilippoPred showed better performance (AUC = 0.908) than the individual predictors SIFT (AUC = 0.756), Polyphen-2 (AUC = 0.788), CADD (AUC = 0.568), REVEL (AUC = 0.548), MetaLR (AUC = 0.751), and AlphMissense (AUC = 0.885). Using high-confidence labelled NAGLU variants, showed that SanfilippoPred has an 85.7% sensitivity threshold. The poor correlation between the Sanfilippo syndrome phenotype and genotype represents a demand for a new tool to classify its missense variants. This study provides a significant tool for preventing the misinterpretation of missense variants of the Sanfilippo syndrome-relevant genes. Finally, it seems that ML-based pathogenicity predictors and Sanfilippo syndrome-specific prediction tools could be feasible and efficient pathogenicity predictors in the future.

PdmIRD: missense variants pathogenicity prediction for inherited retinal diseases in a disease-specific manner

Accurate discrimination of pathogenic and nonpathogenic variation remains an enormous challenge in clinical genetic testing of inherited retinal diseases (IRDs) patients. Computational methods for predicting variant pathogenicity are the main solutions for this dilemma. The majority of the state-of-the-art variant pathogenicity prediction tools disregard the differences in characteristics among different genes and treat all types of mutations equally. Since missense variants are the most common type of variation in the coding region of the human genome, we developed a novel missense mutation pathogenicity prediction tool, named Prediction of Deleterious Missense Mutation for IRDs (PdmIRD) in this study. PdmIRD was tailored for IRDs-related genes and constructed with the conditional random forest model. Population frequencies and a newly available prediction tool were incorporated into PdmIRD to improve the performance of the model. The evaluation of PdmIRD demonstrated its superior performance over nonspecific tools (areas under the curves, 0.984 and 0.910) and an existing eye abnormalities-specific tool (areas under the curves, 0.975 and 0.891). We also demonstrated the submodel that used a smaller gene panel further slightly improved performance. Our study provides evidence that a disease-specific model can enhance the prediction of missense mutation pathogenicity, especially when new and important features are considered. Additionally, this study provides guidance for exploring the characteristics and functions of the mutated proteins in a greater number of Mendelian disorders.

mvPPT: A Highly Efficient and Sensitive Pathogenicity Prediction Tool for Missense Variants

Next-generation sequencing technologies both boost the discovery of variants in the human genome and exacerbate the challenges of pathogenic variant identification. In this study, we developed Pathogenicity Prediction Tool for missense variants (mvPPT), a highly sensitive and accurate missense variant classifier based on gradient boosting. mvPPT adopts high-confidence training sets with a wide spectrum of variant profiles, and extracts three categories of features, including scores from existing prediction tools, frequencies (allele frequencies, amino acid frequencies, and genotype frequencies), and genomic context. Compared with established predictors, mvPPT achieves superior performance in all test sets, regardless of data source. In addition, our study also provides guidance for training set and feature selection strategies, as well as reveals highly relevant features, which may further provide biological insights into variant pathogenicity. mvPPT is freely available at http://www.mvppt.club/.

Exome sequencing in a Swedish family with PMS2 mutation with varying penetrance of colorectal cancer: investigating the presence of genetic risk modifiers in colorectal cancer risk

OBJECTIVE

Lynch syndrome is caused by germline mutations in the mismatch repair (MMR) genes, such as the PMS2 gene, and is characterised by a familial accumulation of colorectal cancer. The penetrance of cancer in PMS2 carriers is still not fully elucidated as a colorectal cancer risk has been shown to vary between PMS2 carriers, suggesting the presence of risk modifiers.

METHODS

Whole exome sequencing was performed in a Swedish family carrying a PMS2 missense mutation [c.2113G>A, p.(Glu705Lys)]. Thirteen genetic sequence variants were further selected and analysed in a case-control study (724 cases and 711 controls).

RESULTS

The most interesting variant was an 18 bp deletion in gene BAG1. BAG1 has been linked to colorectal tumour progression with poor prognosis and is thought to promote colorectal tumour cell survival through increased NF-κB activity.

CONCLUSIONS

We conclude the genetic architecture behind the incomplete penetrance of PMS2 is complicated and must be assessed in a genome wide manner using large families and multifactorial analysis.

A rare CTSC mutation in Papillon-Lefèvre Syndrome results in abolished serine protease activity and reduced NET formation but otherwise normal neutrophil function

Papillon-Lefèvre Syndrome (PLS) is an autosomal recessive monogenic disease caused by loss-of-function mutations in the CTSC gene, thus preventing the synthesis of the protease Cathepsin C (CTSC) in a proteolytically active form. CTSC is responsible for the activation of the pro-forms of the neutrophil serine proteases (NSPs; Elastase, Proteinase 3 and Cathepsin G), suggesting its involvement in a variety of neutrophil functions. In PLS neutrophils, the lack of CTSC protease activity leads to inactivity of the NSPs. Clinically, PLS is characterized by an early, typically pre-pubertal, onset of severe periodontal pathology and palmoplantar hyperkeratosis. However, PLS is not considered an immune deficiency as patients do not typically suffer from recurrent and severe (bacterial and fungal) infections. In this study we investigated an unusual CTSC mutation in two siblings with PLS, a 503A>G substitution in exon 4 of the CTSC gene, expected to result in an amino acid replacement from tyrosine to cysteine at position 168 of the CTSC protein. Both patients bearing this mutation presented with pronounced periodontal pathology. The characteristics and functions of neutrophils from patients homozygous for the 503A>G CTSC mutation were compared to another previously described PLS mutation (755A>T), and a small cohort of healthy volunteers. Neutrophil lysates from patients with the 503A>G substitution lacked CTSC protein and did not display any CTSC or NSP activity, yet neutrophil counts, morphology, priming, chemotaxis, radical production, and regulation of apoptosis were without any overt signs of alteration. However, NET formation upon PMA-stimulation was found to be severely depressed, but not abolished, in PLS neutrophils.

Desmoplakin and periplakin genetically and functionally contribute to eosinophilic esophagitis

Eosinophilic esophagitis (EoE) is a chronic allergic inflammatory disease with a complex underlying genetic etiology. Herein, we conduct whole-exome sequencing of a multigeneration EoE pedigree (discovery set) and 61 additional multiplex families with EoE (replication set). A series of rare, heterozygous, missense variants are identified in the genes encoding the desmosome-associated proteins DSP and PPL in 21% of the multiplex families. Esophageal biopsies from patients with these variants retain dilated intercellular spaces and decrease DSP and PPL expression even during disease remission. These variants affect barrier integrity, cell motility and RhoGTPase activity in esophageal epithelial cells and have increased susceptibility to calpain-14-mediated degradation. An acquired loss of esophageal DSP and PPL is present in non-familial EoE. Taken together, herein, we uncover a pathogenic role for desmosomal dysfunction in EoE, providing a deeper mechanistic understanding of tissue-specific allergic responses.

Opportunities and challenges for the computational interpretation of rare variation in clinically important genes

Genome sequencing is enabling precision medicine-tailoring treatment to the unique constellation of variants in an individual's genome. The impact of recurrent pathogenic variants is often understood, however there is a long tail of rare genetic variants that are uncharacterized. The problem of uncharacterized rare variation is especially acute when it occurs in genes of known clinical importance with functionally consequential variants and associated mechanisms. Variants of uncertain significance (VUSs) in these genes are discovered at a rate that outpaces current ability to classify them with databases of previous cases, experimental evaluation, and computational predictors. Clinicians are thus left without guidance about the significance of variants that may have actionable consequences. Computational prediction of the impact of rare genetic variation is increasingly becoming an important capability. In this paper, we review the technical and ethical challenges of interpreting the function of rare variants in two settings: inborn errors of metabolism in newborns and pharmacogenomics. We propose a framework for a genomic learning healthcare system with an initial focus on early-onset treatable disease in newborns and actionable pharmacogenomics. We argue that (1) a genomic learning healthcare system must allow for continuous collection and assessment of rare variants, (2) emerging machine learning methods will enable algorithms to predict the clinical impact of rare variants on protein function, and (3) ethical considerations must inform the construction and deployment of all rare-variation triage strategies, particularly with respect to health disparities arising from unbalanced ancestry representation.

Alignment-free method for functional annotation of amino acid substitutions: Application on epigenetic factors involved in hematologic malignancies

For the last couple of decades, there has been a significant growth in sequencing data, leading to an extraordinary increase in the number of gene variants. This places a challenge on the bioinformatics research community to develop and improve computational tools for functional annotation of new variants. Genes coding for epigenetic regulators have important roles in cancer pathogenesis and mutations in these genes show great potential as clinical biomarkers, especially in hematologic malignancies. Therefore, we developed a model that specifically focuses on these genes, with an assumption that it would outperform general models in predicting the functional effects of amino acid substitutions. EpiMut is a standalone software that implements a sequence based alignment-free method. We applied a two-step approach for generating sequence based features, relying on the biophysical and biochemical indices of amino acids and the Fourier Transform as a sequence transformation method. For each gene in the dataset, the machine learning algorithm-Naïve Bayes was used for building a model for prediction of the neutral or disease-related status of variants. EpiMut outperformed state-of-the-art tools used for comparison, PolyPhen-2, SIFT and SNAP2. Additionally, EpiMut showed the highest performance on the subset of variants positioned outside conserved functional domains of analysed proteins, which represents an important group of cancer-related variants. These results imply that EpiMut can be applied as a first choice tool in research of the impact of gene variants in epigenetic regulators, especially in the light of the biomarker role in hematologic malignancies. EpiMut is freely available at https://www.vin.bg.ac.rs/180/tools/epimut.php.

Interpretation of Maturity-Onset Diabetes of the Young Genetic Variants Based on American College of Medical Genetics and Genomics Criteria: Machine-Learning Model Development

Background

Maturity-onset diabetes of the young (MODY) is a group of dominantly inherited monogenic diabetes, with HNF4A-MODY, GCK-MODY, and HNF1A-MODY as the three most common forms based on the causal genes. Molecular diagnosis of MODY is important for precise treatment. Although a DNA variant causing MODY can be assessed based on the criteria of the American College of Medical Genetics and Genomics (ACMG) guidelines, gene-specific assessment of disease-causing mutations is important to differentiate among MODY subtypes. As the ACMG criteria were not originally designed for machine-learning algorithms, they are not true independent variables.

Objective

The aim of this study was to develop machine-learning models for interpretation of DNA variants and MODY diagnosis using the ACMG criteria.

Methods

We applied machine-learning models for interpretation of DNA variants in MODY genes defined by the ACMG criteria based on the Human Gene Mutation Database (HGMD) and ClinVar database.

Results

With a machine-learning procedure, we found that the weight matrix of the ACMG criteria was significantly different between the three MODY genes HNF1A, HNF4A, and GCK. The models showed high predictive abilities with accuracy over 95%.

Conclusions

Our results highlight the need for applying different weights of the ACMG criteria in relation to different MODY genes for accurate functional classification. As proof of principle, we applied the ACMG criteria as feature vectors in a machine-learning model and obtained a precision-based result.

REVEL and BayesDel outperform other in silico meta-predictors for clinical variant classification

Many in silico predictors of genetic variant pathogenicity have been previously developed, but there is currently no standard application of these algorithms for variant assessment. Using 4,094 ClinVar-curated missense variants in clinically actionable genes, we evaluated the accuracy and yield of benign and deleterious evidence in 5 in silico meta-predictors, as well as agreement of SIFT and PolyPhen2, and report the derived thresholds for the best performing predictor(s). REVEL and BayesDel outperformed all other meta-predictors (CADD, MetaSVM, Eigen), with higher positive predictive value, comparable negative predictive value, higher yield, and greater overall prediction performance. Agreement of SIFT and PolyPhen2 resulted in slightly higher yield but lower overall prediction performance than REVEL or BayesDel. Our results support the use of gene-level rather than generalized thresholds, when gene-level thresholds can be estimated. Our results also support the use of 2-sided thresholds, which allow for uncertainty, rather than a single, binary cut-point for assigning benign and deleterious evidence. The gene-level 2-sided thresholds we derived for REVEL or BayesDel can be used to assess in silico evidence for missense variants in accordance with current classification guidelines.

Role of deleterious single nucleotide variants in the coding regions of TNFAIP3 for Japanese autoimmune hepatitis with cirrhosis

Autoimmune hepatitis (AIH) is an autoimmune liver disease and cirrhosis is sometimes complicated with AIH at diagnosis, influencing its prognosis. TNFAIP3 gene encodes A20, an inhibitor of nuclear factor-κB pathway, and is a susceptibility gene for autoimmune diseases. We investigated deleterious variants in the coding regions of TNFAIP3 gene of Japanese AIH patients or those with cirrhosis. The deleterious variants in the coding regions of TNFAIP3 gene were analyzed by the cycle sequencing method and the frequencies of deleterious TNFAIP3 alleles of AIH or AIH with cirrhosis were compared with those of Japanese controls. The deleterious alleles in TNFAIP3 were not associated with AIH. A significant association was shown for the deleterious alleles in TNFAIP3 (P = 0.0180, odds ratio (OR) 4.28, 95% confidence interval (CI) 1.53-11.95) with AIH with cirrhosis at presentation. The serum IgM levels in AIH patients with deleterious alleles in TNFAIP3 were tended to be lower than those without (P = 0.0152, Q = 0.1216). The frequency of deleterious alleles in TNFAIP3 was higher in the AIH subset without the DRB1 risk alleles than that with (P = 0.0052, OR 5.10, 95%CI 1.55-16.74). The deleterious alleles in TNFAIP3were associated with AIH with cirrhosis.

Genetic interpretation and clinical translation of minor genes related to Brugada syndrome

Brugada syndrome (BrS) is an inherited arrhythmogenic disease associated with sudden cardiac death. The main gene is SCN5A. Additional variants in 42 other genes have been reported as deleterious, although these variants have not yet received comprehensive pathogenic analysis. Our aim was to clarify the role of all currently reported variants in minor genes associated with BrS. We performed a comprehensive analysis according to the American College of Medical Genetics and Genomics guidelines of published clinical and basic data on all genes (other than SCN5A) related to BrS. Our results identified 133 rare variants potentially associated with BrS. After applying current recommendations, only six variants (4.51%) show a conclusive pathogenic role. All definitively pathogenic variants were located in four genes encoding sodium channels or related proteins: SLMAP, SEMA3A, SCNN1A, and SCN2B. In total, 33.83% of variants in 19 additional genes were potentially pathogenic. Beyond SCN5A, we conclude definitive pathogenic variants associated with BrS in four minor genes. The current list of genes associated with BrS, therefore, should include SCN5A, SLMAP, SEMA3A, SCNN1A, and SCN2B. Comprehensive genetic interpretation and careful clinical translation should be done for all variants currently classified as potentially deleterious for BrS.

Systematic re-evaluation of SCN5A variants associated with Brugada syndrome

BACKGROUND

A large number of SCN5A variants have been reported to underlie Brugada syndrome (BrS). However, the evidence supporting individual variants is highly heterogeneous.

OBJECTIVE

We systematically re-evaluated all SCN5A variants reported in BrS using the 2015 American college of medical genetics and genomics and the association for molecular pathology (ACMG-AMP) guidelines.

METHODS

A PubMed/Embase search was performed to identify all reported SCN5A variants in BrS. Standardized bioinformatic re-analysis (SIFT, PolyPhen, Mutation Taster, Mutation assessor, FATHMM, GERP, PhyloP, and SiPhy) and re-evaluation of frequency in the gnomAD database were performed. Fourteen ACMG-AMP rules were deemed applicable for SCN5A variant analysis.

RESULTS

Four hundred and eighty unique SCN5A variants were identified, the majority of which 425 (88%) were coding variants. One hundred and fifty-six of 425 (37%) variants were classified as pathogenic/likely pathogenic. Two hundred and fifty-eight (60%) were classified as variants of uncertain significance, while a further 11 (3%) were classified as benign/likely benign. When considering the subset of variants that were considered "null" variants separately, 95% fulfilled criteria for pathogenicity/likely pathogenicity. In contrast, only 17% of missense variants fulfilled criteria for pathogenicity/likely pathogenicity. Importantly, however, only 25% of missense variants had available functional data, which was a major score driver for pathogenic classification.

CONCLUSION

Based on contemporary ACMG-AMP guidelines, only a minority of SCN5A variants implicated in BrS fulfill the criteria for pathogenicity or likely pathogenicity.

ClinPred: Prediction Tool to Identify Disease-Relevant Nonsynonymous Single-Nucleotide Variants

Advances in high-throughput DNA sequencing have revolutionized the discovery of variants in the human genome; however, interpreting the phenotypic effects of those variants is still a challenge. While several computational approaches to predict variant impact are available, their accuracy is limited and further improvement is needed. Here, we introduce ClinPred, an efficient tool for identifying disease-relevant nonsynonymous variants. Our predictor incorporates two machine learning algorithms that use existing pathogenicity scores and, notably, benefits from inclusion of normal population allele frequency from the gnomAD database as an input feature. Another major strength of our approach is the use of ClinVar-a rapidly growing database that allows selection of confidently annotated disease-causing variants-as a training set. Compared to other methods, ClinPred showed superior accuracy for predicting pathogenicity, achieving the highest area under the curve (AUC) score and increasing both the specificity and sensitivity in different test datasets. It also obtained the best performance according to various other metrics. Moreover, ClinPred performance remained robust with respect to disease type (cancer or rare disease) and mechanism (gain or loss of function). Importantly, we observed that adding allele frequency as a predictive feature-as opposed to setting fixed allele frequency cutoffs-boosts the performance of prediction. We provide pre-computed ClinPred scores for all possible human missense variants in the exome to facilitate its use by the community.

A Bayesian framework for efficient and accurate variant prediction

There is a growing need to develop variant prediction tools capable of assessing a wide spectrum of evidence. We present a Bayesian framework that involves aggregating pathogenicity data across multiple in silico scores on a gene-by-gene basis and multiple evidence statistics in both quantitative and qualitative forms, and performs 5-tiered variant classification based on the resulting probability credible interval. When evaluated in 1,161 missense variants, our gene-specific in silico model-based meta-predictor yielded an area under the curve (AUC) of 96.0% and outperformed all other in silico predictors. Multifactorial model analysis incorporating all available evidence yielded 99.7% AUC, with 22.8% predicted as variants of uncertain significance (VUS). Use of only 3 auto-computed evidence statistics yielded 98.6% AUC with 56.0% predicted as VUS, which represented sufficient accuracy to rapidly assign a significant portion of VUS to clinically meaningful classifications. Collectively, our findings support the use of this framework to conduct large-scale variant prioritization using in silico predictors followed by variant prediction and classification with a high degree of predictive accuracy.

Cystinuria Associated with Different SLC7A9 Gene Variants in the Cat

Cystinuria is a classical inborn error of metabolism characterized by a selective proximal renal tubular defect affecting cystine, ornithine, lysine, and arginine (COLA) reabsorption, which can lead to uroliths and urinary obstruction. In humans, dogs and mice, cystinuria is caused by variants in one of two genes, SLC3A1 and SLC7A9, which encode the rBAT and bo,+AT subunits of the bo,+ basic amino acid transporter system, respectively. In this study, exons and flanking regions of the SLC3A1 and SLC7A9 genes were sequenced from genomic DNA of cats (Felis catus) with COLAuria and cystine calculi. Relative to the Felis catus-6.2 reference genome sequence, DNA sequences from these affected cats revealed 3 unique homozygous SLC7A9 missense variants: one in exon 5 (p.Asp236Asn) from a non-purpose-bred medium-haired cat, one in exon 7 (p.Val294Glu) in a Maine Coon and a Sphinx cat, and one in exon 10 (p.Thr392Met) from a non-purpose-bred long-haired cat. A genotyping assay subsequently identified another cystinuric domestic medium-haired cat that was homozygous for the variant originally identified in the purebred cats. These missense variants result in deleterious amino acid substitutions of highly conserved residues in the bo,+AT protein. A limited population survey supported that the variants found were likely causative. The remaining 2 sequenced domestic short-haired cats had a heterozygous variant at a splice donor site in intron 10 and a homozygous single nucleotide variant at a branchpoint in intron 11 of SLC7A9, respectively. This study identifies the first SLC7A9 variants causing feline cystinuria and reveals that, as in humans and dogs, this disease is genetically heterogeneous in cats.

The role of functional data in interpreting the effects of genetic variation

Progress in DNA-sequencing technologies has provided a catalogue of millions of DNA variants in the human population, but characterization of the functional effects of these variants has lagged far behind. For example, sequencing of tumor samples is driving an urgent need to classify whether or not mutations seen in cancers affect disease progression or treatment effectiveness or instead are benign. Furthermore, mutations can interact with genetic background and with environmental effects. A new approach, termed deep mutational scanning, has enabled the quantitative assessment of the effects of thousands of mutations in a protein. However, this type of experiment is carried out in model organisms, tissue culture, or in vitro; typically addresses only a single biochemical function of a protein; and is generally performed under a single condition. The current challenge lies in using these functional data to generate useful models for the phenotypic consequences of genetic variation in humans.

Secondary findings and carrier test frequencies in a large multiethnic sample

BACKGROUND

Besides its growing importance in clinical diagnostics and understanding the genetic basis of Mendelian and complex diseases, whole exome sequencing (WES) is a rich source of additional information of potential clinical utility for physicians, patients and their families. We analyzed the frequency and nature of single nucleotide variants (SNVs) considered secondary findings and recessive disease allele carrier status in the exomes of 8554 individuals from a large, randomly sampled cohort study and 2514 patients from a study of presumed Mendelian disease having undergone WES.

METHODS

We used the same sequencing platform and data processing pipeline to analyze all samples and characterized the distributions of reported pathogenic (ClinVar, Human Gene Mutation Database (HGMD)) and predicted deleterious variants in the pre-specified American College of Medical Genetics and Genomics (ACMG) secondary findings and recessive disease genes in different ethnic groups.

RESULTS

In the 56 ACMG secondary findings genes, the average number of predicted deleterious variants per individual was 0.74, and the mean number of ClinVar reported pathogenic variants was 0.06. We observed an average of 10 deleterious and 0.78 ClinVar reported pathogenic variants per individual in 1423 autosomal recessive disease genes. By repeatedly sampling pairs of exomes, 0.5 % of the randomly generated couples were at 25 % risk of having an affected offspring for an autosomal recessive disorder based on the ClinVar variants.

CONCLUSIONS

By investigating reported pathogenic and novel, predicted deleterious variants we estimated the lower and upper limits of the population fraction for which exome sequencing may reveal additional medically relevant information. We suggest that the observed wide range for the lower and upper limits of these frequency numbers will be gradually reduced due to improvement in classification databases and prediction algorithms.