The technology, called STRIPE, enables cost-effective, scalable analysis of full-length RNA molecules to reveal disease-causing genetic variants and provide molecular diagnoses for previously undiagnosed patients

Researchers from Children’s Hospital of Philadelphia (CHOP) developed a new RNA sequencing strategy that can reveal how genetic variants disrupt gene function and improve the diagnosis of rare diseases. In a study published today in the journal Science Advances, the study team demonstrated that this platform could reveal disease-causing genetic variants and provide molecular diagnoses for previously undiagnosed patients, including five individuals whose conditions had remained unresolved after standard testing.

Exome and genome sequencing are widely used methods for identifying genetic variants responsible for rare diseases. However, these approaches have a diagnostic yield of only 20% to 50%, meaning that more than half of patients with suspected rare diseases are unable to obtain a molecular diagnosis. Many genetic variants cause disease by disrupting how RNA molecules are transcribed and processed, meaning their effects cannot always be understood from DNA sequence alone. As a result, researchers and clinicians are increasingly turning to RNA sequencing to better interpret how genetic variants alter gene activity and function and cause disease.

“RNA is a powerful modality for the diagnosis of rare diseases,” said lead senior author Yi Xing, PhD, Associate Chief Scientific Officer for Omics, Technology & Engineering and Francis West Lewis Chair in Computational and Genomic Medicine at CHOP. “By directly observing RNA molecules, we can obtain a more complete picture of how genetic variants alter gene products, in ways that DNA sequencing alone cannot reveal.”

Traditional RNA sequencing methods fragment RNA molecules before sequencing, making it difficult to reconstruct full-length RNA molecules and link disease-associated variants with abnormal RNA processing events across the same molecule. In contrast, long-read RNA sequencing can directly sequence full-length RNA molecules end-to-end, offering the potential to transform RNA-guided interpretation of genetic variants. However, several challenges, including accuracy, cost, and scalability, have limited the widespread use of long-read RNA sequencing for rare diseases to date.

To address these barriers, researchers at CHOP developed STRIPE (Sequencing Targeted RNAs Identifies Pathogenic Events), a targeted long-read RNA sequencing strategy that enables deep sequencing of full-length RNA molecules for any customized disease-specific gene panel.

Overview of STRIPE workflow and quality control

(A) Clinically accessible tissue from a patient is collected for RNA extraction and full-length cDNA synthesis. Resulting cDNA is enriched using TEQUILA probes targeting disease-specific genes and prepared for long-read nanopore sequencing. (B) For each target gene, heterozygous SNVs in TEQUILA-seq reads are used to construct haplotypes, and a realignment strategy is used to partition reads across the haplotypes (left). The resulting read sets are analyzed for extreme patterns in gene expression dosage and splicing by comparison to tissue-matched short-read RNA-seq samples from the GTEx Consortium, as well as the presence of rare deleterious variants based on predicted severity, population allele frequencies in gnomAD, and existing clinical annotations. (C) We sequenced skin fibroblast RNAs from 88 individuals, including 20 unaffected controls and 68 patients with rare disease clinically referred for CDG or PMD. Of the 68 patients, 22 had known genetic causes from prior testing, while the remaining 46 were genetically undiagnosed. (D) Gene abundances based on untargeted long-read RNA-seq and TEQUILA-seq data for fibroblast cell lines CDG-P12 (CDG-466 gene panel) and PMD-C01 (PMD-359 gene panel). Each bar represents one gene, and only the 1000 most abundant genes are shown. On-target rates represent the fraction of transcriptional abundance from target genes. Fold enrichment is calculated by dividing the on-target rate in TEQUILA-seq data by the on-target rate in untargeted data. (E) Comparison of annotated splice junction usage frequencies in target genes between untargeted and TEQUILA-seq data for CDG-P12 (CDG-466 gene panel) and PMD-C01 (PMD-359 gene panel). (F) Principal component analysis (PCA) of annotated splice junction usage frequencies in target genes from the CDG-466 and PMD-359 gene panels in 88 cohort fibroblast samples and GTEx samples for clinically accessible tissues. MAF, minor allele frequency.

STRIPE builds upon prior work at CHOP developing TEQUILA-seq, a scalable and low-cost technology for sequencing full-length RNA molecules.

“TEQUILA-seq was designed to make targeted long-read RNA sequencing cost-effective and scalable,” said co-senior author Lan Lin, PhD, assistant professor of Pathology and Laboratory Medicine at CHOP and developer of the TEQUILA-seq technology. “With an RNA-to-data cost of around $100 per sample, STRIPE enables ultra-deep, full-length RNA sequencing of disease-relevant genes at a scale that is practical for clinical applications.”

To evaluate STRIPE, researchers applied the platform to two groups of rare diseases that are extensively studied at CHOP – congenital disorders of glycosylation (CDG) and primary mitochondrial diseases (PMD). These disease classes include many genetically diverse conditions, with new disease-causing variants continuing to be discovered, making them well suited to assess the effectiveness of STRIPE.

“A major challenge in RNA-guided rare disease diagnostics is that disease-relevant tissues are often difficult to obtain from patients,” said co-senior author Rebecca Ganetzky, MD, an attending physician and clinical geneticist in the Mitochondrial Medicine Program at CHOP. “STRIPE enables high-quality analysis of RNA from clinically accessible tissues such as skin fibroblasts and blood, while still capturing the disease-relevant signals needed to interpret the RNA-level effects of genetic variants.”

“This was a mutually beneficial collaboration in which STRIPE’s new CDG diagnoses could be validated due to the measurable disruption these patients exhibit in glycosylation and thus prove their technology,” said co-senior author Andrew C. Edmondson, MD, PhD, founding Director of the CDG Clinic and an attending physician with the Division of Human Genetics at CHOP. “In turn, CDG patients with high unmet diagnostic needs could be given access to a novel technology after current standard-of-care testing had failed and ultimately receive a molecular diagnosis, ending their diagnostic odyssey and facilitating them access to appropriate clinical care.”

The researchers applied STRIPE to 88 individuals across the two disease groups and healthy controls. The platform accurately re-identified known disease-causing variants and revealed the often complex and sometimes unanticipated consequences of these variants at the RNA level. Importantly, STRIPE clarified the role of previously identified variants with uncertain significance and uncovered new disease-causing variants in five previously undiagnosed patients, enabling clinicians to establish molecular diagnoses that had been elusive.

Since its development, STRIPE has been used to analyze more than 500 patients across multiple clinical programs at CHOP, demonstrating its real-world potential and scalability for rare disease diagnostics.

“By directly revealing how genetic variants disrupt RNA molecules, STRIPE provides a bridge from genetic diagnosis to disease mechanism to targeted therapies,” Xing said. “More broadly, this work reflects a long-standing effort to interpret genetic variants at the level of full-length RNA molecules, and we believe STRIPE can serve as a foundation for RNA-based precision medicine in rare diseases, linking precision diagnostics to precision therapeutics.”

Source – Children’s Hospital of Philadelphia (CHOP)