Frequently used bioinformatics tools overestimate the damaging effect of allelic variants

Accuracy of renovo predictions on variants reclassified over time

BACKGROUND

Interpreting the clinical consequences of genetic variants is the central problem in modern clinical genomics, for both hereditary diseases and oncology. However, clinical validation lags behind the pace of discovery, leading to distressing uncertainty for patients, physicians and researchers. This "interpretation gap" changes over time as evidence accumulates, and variants initially deemed of uncertain (VUS) significance may be subsequently reclassified in pathogenic/benign. We previously developed RENOVO, a random forest-based tool able to predict variant pathogenicity based on publicly available information from GnomAD and dbNFSP, and tested on variants that have changed their classification status over time. Here, we comprehensively evaluated the accuracy of RENOVO predictions on variants that have been reclassified over the last four years.

METHODS

we retrieved 16 retrospective instances of the ClinVar database, every 3 months since March 2020 to March 2024, and analyzed time trends of variant classifications. We identified variants that changed their status over time and compared RENOVO predictions generated in 2020 with the actual reclassifications.

RESULTS

VUS have become the most represented class in ClinVar (44.97% vs. 9.75% (likely) pathogenic and 40,33% (likely) benign). The rate of VUS reclassification is linear and slow compared to the rate of VUS reporting, exponential and currently ~ 30x faster, creating a growing divide between what can be sequenced vs. what can be interpreted. Out of 10,196 VUS variants in January 2020 that have undergone a clinically meaningful reclassification to march 2024, RENOVO correctly classified 82.6% in 2020. In addition, RENOVO correctly identified the majority of the few variants that switched clinically meaningful classes (e.g., from benign to pathogenic and vice versa). We highlight variant classes and clinically relevant genes for which RENOVO provides particularly accurate estimates. In particularly, genes characterized by large prevalence of high- or low-impact variants (e.g., POLE, NOTCH1, FANCM etc.). Suboptimal RENOVO predictions mostly concern genes validated through dedicated consortia (e.g., BRCA1/2), in which RENOVO would anyway have a limited impact.

CONCLUSIONS

Time trend analysis demonstrates that the current model of variant interpretation cannot keep up with variant discovery. Machine learning-based tools like RENOVO confirm high accuracy that can aid in clinical practice and research.

SARS-CoV-2 suppresses IFNβ production mediated by NSP1, 5, 6, 15, ORF6 and ORF7b but does not suppress the effects of added interferon

Type I Interferons (IFN-Is) are a family of cytokines which play a major role in inhibiting viral infection. Resultantly, many viruses have evolved mechanisms in which to evade the IFN-I response. Here we tested the impact of expression of 27 different SARS-CoV-2 genes in relation to their effect on IFN production and activity using three independent experimental methods. We identified six gene products; NSP6, ORF6, ORF7b, NSP1, NSP5 and NSP15, which strongly (>10-fold) blocked MAVS-induced (but not TRIF-induced) IFNβ production. Expression of the first three of these SARS-CoV-2 genes specifically blocked MAVS-induced IFNβ-promoter activity, whereas all six genes induced a collapse in IFNβ mRNA levels, corresponding with suppressed IFNβ protein secretion. Five of these six genes furthermore suppressed MAVS-induced activation of IFNλs, however with no effect on IFNα or IFNγ production. In sharp contrast, SARS-CoV-2 infected cells remained extremely sensitive to anti-viral activity exerted by added IFN-Is. None of the SARS-CoV-2 genes were able to block IFN-I signaling, as demonstrated by robust activation of Interferon Stimulated Genes (ISGs) by added interferon. This, despite the reduced levels of STAT1 and phospho-STAT1, was likely caused by broad translation inhibition mediated by NSP1. Finally, we found that a truncated ORF7b variant that has arisen from a mutant SARS-CoV-2 strain harboring a 382-nucleotide deletion associating with mild disease (Δ382 strain identified in Singapore & Taiwan in 2020) lost its ability to suppress type I and type III IFN production. In summary, our findings support a multi-gene process in which SARS-CoV-2 blocks IFN-production, with ORF7b as a major player, presumably facilitating evasion of host detection during early infection. However, SARS-CoV-2 fails to suppress IFN-I signaling thus providing an opportunity to exploit IFN-Is as potential therapeutic antiviral drugs.

Two cGAS-like receptors induce antiviral immunity in Drosophila

In mammals, cyclic GMP-AMP (cGAMP) synthase (cGAS) produces the cyclic dinucleotide 2'3'-cGAMP in response to cytosolic DNA and this triggers an antiviral immune response. cGAS belongs to a large family of cGAS/DncV-like nucleotidyltransferases that is present in both prokaryotes¹ and eukaryotes^2-5. In bacteria, these enzymes synthesize a range of cyclic oligonucleotides and have recently emerged as important regulators of phage infections^6-8. Here we identify two cGAS-like receptors (cGLRs) in the insect Drosophila melanogaster. We show that cGLR1 and cGLR2 activate Sting- and NF-κB-dependent antiviral immunity in response to infection with RNA or DNA viruses. cGLR1 is activated by double-stranded RNA to produce the cyclic dinucleotide 3'2'-cGAMP, whereas cGLR2 produces a combination of 2'3'-cGAMP and 3'2'-cGAMP in response to an as-yet-unidentified stimulus. Our data establish cGAS as the founding member of a family of receptors that sense different types of nucleic acids and trigger immunity through the production of cyclic dinucleotides beyond 2'3'-cGAMP.

Computational Studies of the Structural Basis of Human RPS19 Mutations Associated With Diamond-Blackfan Anemia

Diamond-Blackfan Anemia (DBA) is an inherited rare disease characterized with severe pure red cell aplasia, and it is caused by the defective ribosome biogenesis stemming from the impairment of ribosomal proteins. Among all DBA-associated ribosomal proteins, RPS19 affects most patients and carries most DBA mutations. Revealing how these mutations lead to the impairment of RPS19 is highly demanded for understanding the pathogenesis of DBA, but a systematic study is currently lacking. In this work, based on the complex structure of human ribosome, we comprehensively studied the structural basis of DBA mutations of RPS19 by using computational methods. Main structure elements and five conserved surface patches involved in RPS19-18S rRNA interaction were identified. We further revealed that DBA mutations would destabilize RPS19 through disrupting the hydrophobic core or breaking the helix, or perturb the RPS19-18S rRNA interaction through destroying hydrogen bonds, introducing steric hindrance effect, or altering surface electrostatic property at the interface. Moreover, we trained a machine-learning model to predict the pathogenicity of all possible RPS19 mutations. Our work has laid a foundation for revealing the pathogenesis of DBA from the structural perspective.

Pathogenic missense protein variants affect different functional pathways and proteomic features than healthy population variants

Missense variants are present amongst the healthy population, but some of them are causative of human diseases. A classification of variants associated with “healthy” or “diseased” states is therefore not always straightforward. A deeper understanding of the nature of missense variants in health and disease, the cellular processes they may affect, and the general molecular principles which underlie these differences is essential to offer mechanistic explanations of the true impact of pathogenic variants. Here, we have formalised a statistical framework which enables robust probabilistic quantification of variant enrichment across full-length proteins, their domains, and 3D structure-defined regions. Using this framework, we validate and extend previously reported trends of variant enrichment in different protein structural regions (surface/core/interface). By examining the association of variant enrichment with available functional pathways and transcriptomic and proteomic (protein half-life, thermal stability, abundance) data, we have mined a rich set of molecular features which distinguish between pathogenic and population variants: Pathogenic variants mainly affect proteins involved in cell proliferation and nucleotide processing and are enriched in more abundant proteins. Additionally, rare population variants display features closer to common than pathogenic variants. We validate the association between these molecular features and variant pathogenicity by comparing against existing in silico variant impact annotations. This study provides molecular details into how different proteins exhibit resilience and/or sensitivity towards missense variants and provides the rationale to prioritise variant-enriched proteins and protein domains for therapeutic targeting and development. The ZoomVar database, which we created for this study, is available at fraternalilab.kcl.ac.uk/ZoomVar. It allows users to programmatically annotate missense variants with protein structural information and to calculate variant enrichment in different protein structural regions.

How do can one improve the classification of genetic variants as harmful or harmless? This study uses a robust statistical analysis to exploit the interplay between protein structure, proteomic measurements and functional pathways to enable better discrimination between missense variants in health and disease.

Variant effect predictions capture some aspects of deep mutational scanning experiments

BACKGROUND

Deep mutational scanning (DMS) studies exploit the mutational landscape of sequence variation by systematically and comprehensively assaying the effect of single amino acid variants (SAVs; also referred to as missense mutations, or non-synonymous Single Nucleotide Variants - missense SNVs or nsSNVs) for particular proteins. We assembled SAV annotations from 22 different DMS experiments and normalized the effect scores to evaluate variant effect prediction methods. Three trained on traditional variant effect data (PolyPhen-2, SIFT, SNAP2), a regression method optimized on DMS data (Envision), and a naïve prediction using conservation information from homologs.

RESULTS

On a set of 32,981 SAVs, all methods captured some aspects of the experimental effect scores, albeit not the same. Traditional methods such as SNAP2 correlated slightly more with measurements and better classified binary states (effect or neutral). Envision appeared to better estimate the precise degree of effect. Most surprising was that the simple naïve conservation approach using PSI-BLAST in many cases outperformed other methods. All methods captured beneficial effects (gain-of-function) significantly worse than deleterious (loss-of-function). For the few proteins with multiple independent experimental measurements, experiments differed substantially, but agreed more with each other than with predictions.

CONCLUSIONS

DMS provides a new powerful experimental means of understanding the dynamics of the protein sequence space. As always, promising new beginnings have to overcome challenges. While our results demonstrated that DMS will be crucial to improve variant effect prediction methods, data diversity hindered simplification and generalization.

Identification of novel FBN1 variations implicated in congenital scoliosis

Congenital scoliosis (CS) is a form of scoliosis caused by congenital vertebral malformations. Genetic predisposition has been demonstrated in CS. We previously reported that TBX6 loss-of-function causes CS in a compound heterozygous model; however, this model can explain only 10% of CS. Many monogenic and polygenic CS genes remain to be elucidated. In this study, we analyzed exome sequencing (ES) data of 615 Chinese CS from the Deciphering Disorders Involving Scoliosis and COmorbidities (DISCO) project. Cosegregation studies for 103 familial CS identified a novel heterozygous nonsense variant, c.2649G>A (p.Trp883Ter) in FBN1. The association between FBN1 and CS was then analyzed by extracting FBN1 variants from ES data of 574 sporadic CS and 828 controls; 30 novel variants were identified and prioritized for further analyses. A mutational burden test showed that the deleterious FBN1 variants were significantly enriched in CS subjects (OR = 3.9, P = 0.03 by Fisher’s exact test). One missense variant, c.2613A>C (p.Leu871Phe) was recurrent in two unrelated CS subjects, and in vitro functional experiments for the variant suggest that FBN1 may contribute to CS by upregulating the transforming growth factor beta (TGF-β) signaling. Our study expanded the phenotypic spectrum of FBN1, and provided nove insights into the genetic etiology of CS.

Host Genetics, Innate Immune Responses, and Cellular Death Pathways in Poliomyelitis Patients

Purpose

Poliovirus (PV) is one of the most studied viruses. Despite efforts to understand PV infection within the host, fundamental questions remain unanswered. These include the mechanisms determining the progression to viremia, the pathogenesis of neuronal infection and paralysis in only a minority of patients. Because of the rare disease phenotype of paralytic poliomyelitis (PPM), we hypothesize that a genetic etiology may contribute to the disease course and outcome.

Methods

We used whole-exome sequencing (WES) to investigate the genetic profile of 18 patients with PPM. Functional analyses were performed on peripheral blood mononuclear cells (PBMCs) and monocyte-derived macrophages (MdMs).

Results

We identified rare variants in host genes involved in interferon signaling, viral replication, apoptosis, and autophagy. Upon PV infection of MdMs, we observed a tendency toward increased viral burden in patients compared to controls, suggesting reduced control of PV infection. In MdMs from patients, the IFNβ response correlated with the viral burden.

Conclusion

We suggest that genetic variants in innate immune defenses and cell death pathways contribute to the clinical presentation of PV infection. Importantly, this study is the first to uncover the genetic profile in patients with PPM combined with investigations of immune responses and viral burden in primary cells.

Identification and characterization of rare toll-like receptor 3 variants in patients with autoimmune Addison's disease

Autoimmune Addison's disease (AAD) is a classic organ-specific autoimmune disease characterized by an immune-mediated attack on the adrenal cortex. As most autoimmune diseases, AAD is believed to be caused by a combination of genetic and environmental factors, and probably interactions between the two. Persistent viral infections have been suggested to play a triggering role, by invoking inflammation and autoimmune destruction. The inability of clearing infections can be due to aberrations in innate immunity, including mutations in genes involved in the recognition of conserved microbial patterns. In a whole exome sequencing study of anonymized AAD patients, we discovered several rare variants predicted to be damaging in the gene encoding Toll-like receptor 3 (TLR3). TLR3 recognizes double stranded RNAs, and is therefore a major factor in antiviral defense. We here report the occurrence and functional characterization of five rare missense variants in TLR3 of patients with AAD. Most of these variants occurred together with a common TLR3 variant that has been associated with a wide range of immunopathologies. The biological implications of these variants on TLR3 function were evaluated in a cell-based assay, revealing a partial loss-of-function effect of three of the rare variants. In addition, rare mutations in other members of the TLR3-interferon (IFN) signaling pathway were detected in the AAD patients. Together, these findings indicate a potential role for TLR3 and downstream signaling proteins in the pathogenesis in a subset of AAD patients.

Opportunities and Challenges for Genetic Studies of End-Stage Renal Disease in Canada

Purpose of review:

Genetic testing can improve diagnostic precision in some patients with end-stage renal disease (ESRD) providing the potential for targeted therapy and improved patient outcomes. We sought to describe the genetic architecture of ESRD and Canadian data sources available for further genetic investigation into ESRD.

Sources of information:

We performed PubMed searches of English, peer-reviewed articles using keywords “chronic kidney disease,” “ESRD,” “genetics,” “sequencing,” and “administrative databases,” and searched for nephrology-related Mendelian diseases on the Online Mendelian Inheritance in Man database.

Methods:

In this narrative review, we discuss our evolving understanding of the genetic architecture of kidney disease and ESRD, the risks and benefits of using genetic data to help diagnose and manage patients with ESRD, existing public Canadian biobanks and databases, and a vision for future genetic studies of ESRD in Canada.

Key findings:

ESRD has a polygenic architecture including rare Mendelian mutations and common small effect genetic polymorphism contributors. Genetic testing will improve diagnostic accuracy and contribute to a precision medicine approach in nephrology. However, the risk and benefits of genetic testing needs to be considered from an individual and societal perspective, and further research is required. Merging existing health data, linking biobanks and administrative databases, and forming Canadian collaborations hold great potential for genetic research into ESRD. Large sample sizes are necessary to perform the suitably powered investigations required to bring this vision to reality.

Limitations:

This is a narrative review of the literature discussing future directions and opportunities. It reflects the views and academic biases of the authors.

Implications:

National collaborations will be required to obtain sample sizes required for impactful, robust research. Merging established datasets may be one approach to obtain adequate samples. Patient education and engagement will improve the value of knowledge gained.