Increasing the throughput of variant classification with multiplexed prime editing and single cell long-read technologies

The National Institutes of Health National Human Genome Research Institute awarded Northwestern University $100,228 under Project Grant F31HG014981 on November 1, 2026, to develop computational and experimental tools for high-throughput variant classification in genetic medicine.

The recipient will advance multiplexed assays of variant effect (MAVEs) to resolve variants of uncertain significance (VUS)—DNA alterations that cannot be definitively classified as benign or pathogenic—which currently account for approximately 80 percent of missense variants in the ClinVar database. Northwestern will develop a computational pipeline to design prime editing guide RNA libraries with high predicted editing efficiency, an optimized cell line for editing, and a fluorescence-activated cell sorting approach to enrich for edited cells. Using CHD2 as a model gene, the work will develop techniques applicable to hundreds of clinically relevant genes. The research addresses two key limitations in current MAVEs: the difficulty of introducing a comprehensive set of variants into cells at scale and the historical reliance on limited functional assays such as proliferation or survival-based assays.

Work is performed in Chicago, Illinois with a period of performance through May 31, 2028. The award falls under the Human Genome Research program (CFDA 93.172), which supports development of genomic resources, technologies, and methods in basic and functional genomics applicable to biomedical research and the advancement of genomics in medical science and clinical care.

Abstract

In genetic medicine, variants of uncertain significance (VUS) have posed a significant challenge. These variants, identified through genetic testing, are alterations in one's DNA sequence that cannot be definitively classified as benign or pathogenic. Approximately 80% of missense variants in the clinvar database are classified as VUS. In clinical care, a VUS result frequently prevents clinicians from providing definitive diagnoses, limits treatment options, and causes uncertainty for patients and their families. Resolving VUS is thus essential to improve patient care in genetic medicine. Multiplexed assays of variant effect (MAVES) offer a high-throughput solution to VUS resolution by assessing variant effect for thousands of variants in vitro simultaneously. However, there are two limitations to realizing the potential of MAVES to resolve VUS at scale. First, IT is challenging to introduce a comprehensive set of variants into cells at scale, and second, MAVES have historically relied on a small set of function assays, such as proliferation or survival-based assays, limiting the number of genes that can be investigated. This proposal aims to address both of these limitations, using CHD2 as a model gene, but developing techniques that could be applied to hundreds of clinically relevant genes. We will first advance high-throughput endogenous genome editing through the development of a suite of computational and experimental tools for prime editing. While prime editing offers a precise and versatile approach, IT is plagued by low editing efficiencies. To address these shortcomings, we will develop a computational pipeline to design prime editing guide RNA (PEGRNA) libraries with high predicted editing efficiency, an optimized cell line for editing, and a facs-based approach to enrich for edited cells. We expect that such tools will boost prime editing efficiency for multiplexed cell libraries and thereby increase the efficiency of MAVES. We will also address the limited set of functional assays for MAVES. Here we propose scaver-SEQ: a widely applicable and scalable single-cell assay of variant effect using RNA-sequencing. This technique will combine transcript capture and long-read technology in single cells to identify each cell's introduced variant and link IT to its corresponding transcriptome in a single-tube workflow. By using the transcriptome as a functional readout, we will revolutionize variant classification in disorders associated with robust transcriptional effects, such as those caused by pathogenic variants in regulators of gene expression. This approach could help resolve the over 500,000 VUS that exist in gene expression regulators, which comprise 31% of all VUS in clinvar. By developing generalizable methods to characterize VUS, we aim to contribute to the nhgri's mission of understanding the impact of genomic variation on human health and simultaneously help deliver answers to families facing genetic testing results reporting VUS.