Genomics and Rare Diseases

This paper introduces ConSpliceML, a machine learning model that improves the identification of deleterious splicing variants by integrating genetic constraint and alternative splicing predictions.

The paper systematically analyzes chrX genes to predict disease associations, identifying 235 genes likely involved in disorders and reporting damaging variants in CDK16 and TRPC5 in patients with intellectual disability or autism.

The paper uses prior ClinVar variant classifications to inform ACMG/AMP variant interpretation by identifying SNVs with similar functional consequences to pathogenic variants.

The paper evaluates nine in silico pathogenicity prediction tools for small in-frame indels using a dataset from gnomAD, ClinVar, and DDD, finding VEST-indel performs best with AUCs of 0.93 and 0.87.

The paper reports three novel VHL germline mutations in Iranian RCH patients and analyzes their impact on pVHL structure and function using in silico methods.

PhenoScore quantifies phenotypic similarity using AI and HPO data to improve genetic variant interpretation and genotype-phenotype studies.

SUsPECT improves variant annotation by using custom long-read transcriptomes to better predict variant effects in clinical settings.

This study performs a meta-analysis of genotype-phenotype associations in Alström syndrome, finding that mutations in exon 10 are linked to higher liver disease prevalence.

The paper investigates Gly mutations in Mendelian disorders using ClinVar data, finding that bulky hydrophobic amino acid substitutions reduce protein stability and that Gly mutations may create binding pockets targetable for drug development.

The paper estimates penetrance of monogenic diabetes variants across different ascertainment contexts, finding age-related variability for HNF1A/HNF4A variants but consistent penetrance for GCK variants.

The paper introduces Cancermuts, a Python tool for prioritizing missense cancer variants, and validates its predictions on AMBRA1 mutations in melanoma.

FAVOR provides an online resource and tool for functional annotation of variants from whole genome sequencing studies to aid in disease gene prioritization.

E-SNPs&GO improves pathogenic variant annotation by embedding protein sequences and function, achieving a MCC score of 0.71 on a blind test set.

The paper identifies novel Mendelian genes by analyzing mouse embryo lethality and linking them to human disease phenotypes and genetic variants.

The paper develops a system that integrates variant quality control into prioritization to improve the accuracy of identifying disease-causing variants in rare disorders.

The paper adapts ACMG/AMP variant curation guidelines for the myocilin gene, leading to changes in variant classification for 38% of tested variants.

This study reviews the cost and clinical utility of whole exome sequencing (WES) and whole genome sequencing (WGS) in pediatric genetic disease diagnosis, finding limited evidence of WGS superiority over WES in changing clinical management.

This paper introduces MetaRNN and MetaRNN-indel, deep learning models that differentiate rare pathogenic from benign variants, outperforming existing methods and providing interpretable scores.

The paper investigates tubulin mutations' molecular mechanisms and their phenotypic effects, finding that current computational methods poorly predict their pathogenicity due to dominant-negative and gain-of-function mechanisms.

The paper identifies a novel syndrome caused by the gain-of-function variant p.Glu1099Lys in NSD2, leading to distinct clinical features compared to loss-of-function variants.

The paper introduces a variant catalogue pipeline to generate high-quality variant libraries from whole genome sequences, applied to 100 human genomes to improve variant cataloguing for diverse populations.

This study shows that pre-trained protein language models can predict mutation pathogenicity and patient survival in cancer, outperforming existing methods.

The paper demonstrates that recurrent genomic deletions mediated by nonallelic homologous recombination drive recessive disease burden and identifies new disease-contributing variants through exome reanalysis.

The paper develops a machine learning framework that uses phenotypic similarity to predict the functional effects of missense variants in voltage-gated sodium channels, achieving high accuracy and robustness.

The paper introduces HapICE, a method to improve variant detection in SBDS using NGS, achieving higher diagnostic accuracy than conventional methods.

This paper evaluates exome sequencing as an alternative to chromosomal microarray for detecting copy number variants in rare disorders, showing improved diagnostic yield in prospective cases.

GenOtoScope automates ACMG/AMP variant classification for congenital hearing loss, achieving higher accuracy and precision than existing tools.

This study evaluates if low-coverage genome sequencing can detect clinically relevant copy-number and mtDNA variants, finding that 2x coverage is sufficient for most cases.

The paper estimates the prevalence of putatively pathogenic leptin variants in the gnomAD database, finding low frequencies with higher risk in South Asian populations.

RAVA-FIRST improves rare variant association testing in whole-genome data by using functionally-informed testing units and filtering, identifying a novel intergenic region linked to venous thromboembolism.