Biology subjects

Khazeeva, G.

Publications and source records attributed to Khazeeva, G..

3 recordsLinked to original sources

bioRxiv2026

Deep-intronic variants in USH2A: the importance of model selection in variant interpretation and pathogenicity assessment

Interpreting splice-altering USH2A variants is challenging when these result in the alteration of splicing in only a fraction of transcripts. We investigated two deep-intronic USH2A variants from a single case, the inherited variant c.5573-834A>G and the de novo variant c.11047+345A>G, using a minigene splice assay and transcript analysis in patient-derived photoreceptor precursor cells and nasal epithelial cells. The minigene assay revealed the inclusion of a pseudoexon in the presence of the c.11047+345A>G variant, but did not reproduce previously reported aberrant splicing for c.5573-834A>G. In patient-derived cells, c.5573-834A>G resulted in the inclusion of a pseudoexon in part of the transcripts, whereas c.11047+345A>G did not result in detectable levels of aberrantly spliced transcripts. To assess the impact of residual wildtype transcripts on retinal function, we used the previously developed humanized ush2aPE40+M1 knock-in zebrafish model which allows the modeling of the impact of mixed wildtype and pseudoexon-containing ush2a transcripts. This model demonstrated that detectable levels of usherin expression are maintained even when only a fraction of wildtype transcripts lacking pseudoexon inclusion remain. Collectively, these findings advocate for a more nuanced classification of partially splice-altering USH2A variants. The mere presence of alternative transcripts should not be considered sufficient evidence for pathogenicity. Within this framework, data on the c.5573-834A>G variant do not support a (likely) pathogenic classification, and findings related to the c.11047+345A>G variant are consistent with a likely benign classification.

genetics↗

bioRxiv2024

Reproductive and cognitive effects in carriers of recessive pathogenic variants

The genetic landscape of human Mendelian diseases is shaped by mutation and selection. Selection is mediated by phenotypic effects which interfere with health and reproductive success. Although selection on heterozygotes is well-established in autosomal dominant disorders, convincing evidence for selection in carriers of pathogenic variants associated with recessive conditions is limited, with only a few specific cases documented. We studied heterozygous pathogenic variants in 1,929 genes, which cause recessive diseases when bi-allelic, in a cohort of 378,751 unrelated European individuals from the UK Biobank1. We assessed the impact of these pathogenic variants on reproductive success. We find evidence for fitness effects in heterozygous carriers for recessive genes, especially for variants in constrained genes across a broad range of diseases. Our data suggest reproductive effects at the population level, and hence natural selection, for autosomal recessive disease variants. We further show that variants in genes that underlie intellectual disability are associated with reduced cognition measures in carriers. In concordance with this, we observe an altered genetic landscape, characterized by a threefold reduction in the calculated frequency of biallelic intellectual disability in the population relative to other recessive disorders. The existence of phenotypic and selective effects of pathogenic variants in constrained recessive genes is consistent with a gradient of heterozygote effects, rather than a strict dominant-recessive dichotomy2.

genetics↗

bioRxiv2021

DeNovoCNN: A deep learning approach to de novo variant calling in next generation sequencing data

De novo mutations (DNMs) are an important cause of genetic disorders. The accurate identification of DNMs from sequencing data is therefore fundamental to rare disease research and diagnostics. Unfortunately, identifying reliable DNMs remains a major challenge due to sequence errors, uneven coverage, and mapping artifacts. Here, we developed a deep convolutional neural network (CNN) DNM caller (DeNovoCNN), that encodes the alignment of sequence reads for a trio as 160x164 resolution images. DeNovoCNN was trained on DNMs of 5,616 whole exome sequencing (WES) trios achieving total 96.74% recall and 96.55% precision on the test dataset. We find that DeNovoCNN has increased recall/sensitivity and precision compared to existing DNM calling approaches (GATK, DeNovoGear, DeepTrio, Samtools) based on the Genome in a Bottle reference dataset and independent WES and WGS trios. Validations of DNMs based on Sanger and PacBio HiFi sequencing confirm that DeNovoCNN outperforms existing methods. Most importantly, our results suggest that DeNovoCNN is likely robust against different exome sequencing and analyses approaches, thereby allowing the application on other datasets. DeNovoCNN is freely available as a Docker container and can be run on existing alignment (BAM/CRAM) and variant calling (VCF) files from WES and WGS without a need for variant recalling.

bioinformatics↗