Homozygous synonymous FAM111A variant underlies an autosomal recessive form of Kenny-Caffey syndrome

Journal of Human Genetics, Год журнала: 2024, Номер unknown

Опубликована: Ноя. 6, 2024

Abstract FAM111A (family with sequence similarity 111 member A) is a serine protease and removes covalent DNA-protein cross-links during DNA replication. Heterozygous gain-of-function variants in cause skeletal dysplasias, such as the perinatal lethal osteocraniostenosis milder Kenny-Caffey syndrome (KCS). We report two siblings born to consanguineous parents dysmorphic craniofacial features, postnatal growth retardation, ophthalmologic manifestations, hair nail anomalies, abnormalities thickened cortex stenosis of medullary cavity long bones suggestive KCS. Using exome sequencing, homozygous synonymous variant, NM_001312909.2:c.81 G > A; p.Pro27=, that affects last base exon predicted alter pre-mRNA splicing, was identified both siblings. aberrantly spliced transcripts, reduced mRNA levels, near-complete absence protein fibroblasts patients. After treatment patient control different concentrations camptothecin induces cross-links, we observed tendency towards proportion metabolically active cells compared fibroblasts. However, under these culture conditions, did not find consistent statistically significant differences cell cycle progression apoptotic death between cells. Our findings show deficiency underlies an autosomal recessive form -related Based on our results published data, hypothesize loss hyperactivity, for patient-variant proteins, may converge similar pathomechanism underlying dysplasias.

Язык: Английский

AlphaRING: Missense Variant Classification via Protein Modelling and Residue Interaction Network Analysis

bioRxiv (Cold Spring Harbor Laboratory), Год журнала: 2024, Номер unknown

Опубликована: Ноя. 15, 2024

Abstract Recent advancements in protein modelling, particularly with AlphaFold, have opened new avenues for assessing the structural impacts of missense variants on function. Tools like AlphaMissense leverage AlphaFold2 to predict variant pathogenicity state-of-the-art performance. However, is deep learning-based, and its process opaque does not explicitly impact based additional biophysical properties. Additionally, publicly available, limiting accessibility. Here, we present AlphaRING, a available tool that leverages captures real physico-chemical properties residues both characterised uncharacterised variants. AlphaRING models wild-type proteins using identifies atomic-level non-covalent interactions via residue interaction network analysis RING4. Unlike learning models, our scoring mutated residues, providing interpretable insights into mutation effects. We developed novel formulas each bond type (hydrogen, ionic, π-cation, π-π stacking, π-hydrogen), incorporating bond-specific energy geometry values calculate unique weighting scores per residue. The absolute log 2 fold change between yields non-negative score—the higher score, greater predicted pathogenicity. Evaluating over 1,300 human gold-standard curated by ClinVar expert panels, results show effective distinction benign pathogenic variants, receiver operating characteristic area under curve 0.79. This strong performance suggests reliable, open-source suitable high-throughput predictions large-scale characterisation, accessible non-experts. By enabling detailed exploration relationship genetic variation structure, addresses key limitations previous paves way broader applications research precision medicine.

Язык: Английский

Leveraging Protein Language Models to Identify Complex Trait Associations with Previously Inaccessible Classes of Functional Rare Variants

Опубликована: Янв. 1, 2024

Язык: Английский