Biology subjects

Kwint, M.

Publications and source records attributed to Kwint, M..

4 recordsLinked to original sources

bioRxiv2026

Mapping NMD-coupled alternative splicing in iPSC-derived brain cells: a resource for therapeutic discovery

Therapeutic options for neurodevelopmental disorders (NDD) are expanding. Targeting alternative splicing (AS) events linked to nonsense-mediated decay (NMD) offers a promising way to boost gene expression in haploinsufficiency disorders. However, naturally occurring NMD-coupled AS (NMD AS) events in brain cells remain poorly characterized. Here, we integrate long- and short-read RNA sequencing of NMD-inhibited induced pluripotent stem cell-derived excitatory neurons, astrocytes, and microglia to map and prioritize NMD AS events most suitable for therapeutic intervention. We developed an optimized prediction framework and provide an open access, queryable, database cataloging the existence and abundance of NMD AS events across these cell types. Querying this resource, we identified 936 NMD-sensitive AS events in 250 autosomal-dominant NDD genes and nominate 60 NMD AS events in haploinsufficient genes underlying 42 NDDs that are abundant in at least one cell type and thus represent promising therapeutic targets. These included a previously targeted NMD AS event in SCN1A, and known NMD AS events in CHD2, EZH2, and NR4A2, for which we confirm high cell type-specific abundance, as well as an newly identified abundant event in PHIP. Beyond NDD genes, we identify 1,817 differentially spliced AS events including NMD AS events, highlighting the potential functional role of (NMD) AS within brain cell-specific regulatory programs. This framework and resource enables systematic discovery and prioritization of therapeutically targetable NMD AS events and establish a cell-type resolved atlas to guide splice-modulating strategies in NDDs.

genetics↗

bioRxiv2024

Pushing the limits of single molecule transcript sequencing to uncover the largest disease-associated transcript isoforms in the human neural retina

Sequencing technologies have long limited the comprehensive investigation of large transcripts associated with inherited retinal diseases (IRDs) like Usher syndrome, which involves 11 associated genes with transcripts up to 19.6 kb. To address this, we used PacBio long-read mRNA isoform sequencing (Iso-Seq) following standard library preparation and an optimized workflow to enrich for long transcripts in the human neural retina. While our workflow achieved sequencing of transcripts up to 15 kb, this was insufficient for Usher syndrome-associated genes USH2A and ADGRV1, with transcripts of 18.9 kb and 19.6 kb, respectively. To overcome this, we employed the Samplix Xdrop System for indirect target enrichment of cDNA, a technique typically used for genomic DNA capture. This method facilitated the successful capture and sequencing of ADGRV1 transcripts as well as the full-length 18.9 kb USH2A transcripts. By combining algorithmic analysis with detailed manual curation of sequenced reads, we identified novel isoforms and alternative splicing events across the 11 Usher syndrome-associated genes, with implications for diagnostics and therapy development. Our findings demonstrate the Xdrop systems adaptability for cDNA capture and the advantages of integrating computational and manual transcript analyses. The full neural retina sequencing dataset is available via EGA under identifier EGAD50000000720.

genetics↗

bioRxiv2024

A proteogenomic atlas of the human neural retina

The human neural retina is a complex tissue with abundant alternative splicing and more than 10% of genetic variants linked to inherited retinal diseases (IRDs) alter splicing. Traditional short-read RNA-sequencing methods have been used for understanding retina-specific splicing but have limitations in detailing transcript isoforms. To address this, we generated a proteogenomic atlas that combines PacBio long-read RNA-sequencing data with mass spectrometry and whole genome sequencing data of three healthy human neural retina samples. We identified nearly 60,000 transcript isoforms, of which approximately one-third are novel. Additionally, ten novel peptides confirmed novel transcript isoforms. For instance, we identified a novel IMPDH1 isoform with a novel combination of known exons that is supported by peptide evidence. Our research underscores the potential of in-depth tissue-specific transcriptomic analysis to enhance our grasp of tissue-specific alternative splicing. The data underlying the proteogenomic atlas are available via EGA with identifier EGAD50000000101, via ProteomeXchange with identifier PXD045187, and accessible through the UCSC genome browser. O_FIG O_LINKSMALLFIG WIDTH=200 HEIGHT=149 SRC="FIGDIR/small/595273v1_ufig1.gif" ALT="Figure 1"> View larger version (30K): org.highwire.dtl.DTLVardef@1e0810org.highwire.dtl.DTLVardef@1a382b6org.highwire.dtl.DTLVardef@1da7d9eorg.highwire.dtl.DTLVardef@16d29a3_HPS_FORMAT_FIGEXP M_FIG C_FIG

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↗