Investigating the diagnostic yield of long read sequencing in a selected cohort of families with undiagnosed disease
Laboratory Genetics and Genomics
-
Primary Categories:
- Laboratory Genetics
-
Secondary Categories:
- Laboratory Genetics
Introduction:
The University of Utah launched the Penelope Program for rare and undiagnosed diseases in 2015 and was subsequently designated a clinical site for the Undiagnosed Diseases Network (UDN) in 2018. The goal of both programs is to solve some of the most complex medical cases utilizing a team of experienced clinicians, molecular geneticists, data scientists and researchers, spread across multiple specialties. The combined enrolment for both programs as of November 2024 is 195 patients with 66 patients (34%) having received a diagnosis thus far. This unfortunately means that 129 patients (66%) remain molecularly undiagnosed, despite extensive work-up via a variety of techniques, including cytogenetic (e.g. karyotype, microarray), molecular (e.g. FMR1 testing), short-read (SR) sequencing (e.g., whole genome sequencing (WGS), RNA sequencing) and additional testing specifically tailored to each patient.
Current clinical laboratory techniques cannot comprehensively evaluate all disease-causing aberrations. Furthermore, limitations in current SR sequencing technologies mean that all areas of the genome are not adequately assessed (e.g. repetitive elements). Long-read (LR) sequencing technologies have the potential to address many of these shortcomings and to combine testing into one unified platform. This project seeks to assess the clinical utility of Oxford Nanopore Technologies (ONT) LRWGS in this undiagnosed cohort of patients.
Methods:
We reviewed the 129 cases that remain undiagnosed and selected 22, as a pilot cohort, where there remained a strong suspicion for an underlying genetic condition. 7 of these cases originated from the Penelope Program and 15 were from the UDN. For 5 cases, high quality DNA was available, and we opted to use this previously banked material for LR sequencing. This also allowed us to establish the workflow and to test the pipeline for this project. For all other cases, fresh blood samples were obtained, and high molecular weight DNA was extracted. Sequencing was performed on the ONT PromethION platform, with a targeted sequencing depth of 30x. The raw data was processed through ONTs bioinformatic EPI2ME pipeline, for base calling, primary alignment and secondary variant calling. To identify genetic variants that might explain the patient’s presentation, tertiary analysis using phenotype-driven variant prioritization was performed. Variants of interest were subsequently evaluated by the full Penelope/University of Utah UDN team.
Results:
At the time of abstract submission, we have analyzed 5 of the 22 cases. In one of these cases, we identified a de novo 26 bp deletion in the IRF2BPL gene, leading to a pre-mature termination of the IRF2BPL transcript. Due to the presence of a repetitive Poly Gln/Ala tracks, sequencing of this region was problematic for SR technologies, but was readily detected using LR sequencing. Heterozygous nonsense or frameshift variants in the IRF2BPL gene are associated with neurodevelopmental disorder with regression, abnormal movements, loss of speech, and seizures (NEDAMSS; MIM: 618088), and this 26 bp deletion is consistent with this mechanism. Additionally, the patient’s phenotype was consistent with this diagnosis. Diagnostic variants have not yet been identified in the 4 other sequenced cases. We are preparing to sequence the remaining 17 cases, and an up-to-date comprehensive analysis of the whole cohort will be presented at the ACMG Annual Clinical Genetics Meeting.
Conclusion:
LRWGS is reported to increase diagnostic yield by ~10% over SR sequencing technologies. Therefore, from our undiagnosed disease cohort of 22 patients, we anticipate diagnosing 2-3 cases. To this point, we have diagnosed 1 patient out of the first 5 cases. We anticipate diagnosing an additional 1-2 patients from the remaining cases in our cohort. This data would be consistent with previously reported findings on the increased diagnostic yield of LRWGS, which is especially important in these rare and undiagnosed cases.
The University of Utah launched the Penelope Program for rare and undiagnosed diseases in 2015 and was subsequently designated a clinical site for the Undiagnosed Diseases Network (UDN) in 2018. The goal of both programs is to solve some of the most complex medical cases utilizing a team of experienced clinicians, molecular geneticists, data scientists and researchers, spread across multiple specialties. The combined enrolment for both programs as of November 2024 is 195 patients with 66 patients (34%) having received a diagnosis thus far. This unfortunately means that 129 patients (66%) remain molecularly undiagnosed, despite extensive work-up via a variety of techniques, including cytogenetic (e.g. karyotype, microarray), molecular (e.g. FMR1 testing), short-read (SR) sequencing (e.g., whole genome sequencing (WGS), RNA sequencing) and additional testing specifically tailored to each patient.
Current clinical laboratory techniques cannot comprehensively evaluate all disease-causing aberrations. Furthermore, limitations in current SR sequencing technologies mean that all areas of the genome are not adequately assessed (e.g. repetitive elements). Long-read (LR) sequencing technologies have the potential to address many of these shortcomings and to combine testing into one unified platform. This project seeks to assess the clinical utility of Oxford Nanopore Technologies (ONT) LRWGS in this undiagnosed cohort of patients.
Methods:
We reviewed the 129 cases that remain undiagnosed and selected 22, as a pilot cohort, where there remained a strong suspicion for an underlying genetic condition. 7 of these cases originated from the Penelope Program and 15 were from the UDN. For 5 cases, high quality DNA was available, and we opted to use this previously banked material for LR sequencing. This also allowed us to establish the workflow and to test the pipeline for this project. For all other cases, fresh blood samples were obtained, and high molecular weight DNA was extracted. Sequencing was performed on the ONT PromethION platform, with a targeted sequencing depth of 30x. The raw data was processed through ONTs bioinformatic EPI2ME pipeline, for base calling, primary alignment and secondary variant calling. To identify genetic variants that might explain the patient’s presentation, tertiary analysis using phenotype-driven variant prioritization was performed. Variants of interest were subsequently evaluated by the full Penelope/University of Utah UDN team.
Results:
At the time of abstract submission, we have analyzed 5 of the 22 cases. In one of these cases, we identified a de novo 26 bp deletion in the IRF2BPL gene, leading to a pre-mature termination of the IRF2BPL transcript. Due to the presence of a repetitive Poly Gln/Ala tracks, sequencing of this region was problematic for SR technologies, but was readily detected using LR sequencing. Heterozygous nonsense or frameshift variants in the IRF2BPL gene are associated with neurodevelopmental disorder with regression, abnormal movements, loss of speech, and seizures (NEDAMSS; MIM: 618088), and this 26 bp deletion is consistent with this mechanism. Additionally, the patient’s phenotype was consistent with this diagnosis. Diagnostic variants have not yet been identified in the 4 other sequenced cases. We are preparing to sequence the remaining 17 cases, and an up-to-date comprehensive analysis of the whole cohort will be presented at the ACMG Annual Clinical Genetics Meeting.
Conclusion:
LRWGS is reported to increase diagnostic yield by ~10% over SR sequencing technologies. Therefore, from our undiagnosed disease cohort of 22 patients, we anticipate diagnosing 2-3 cases. To this point, we have diagnosed 1 patient out of the first 5 cases. We anticipate diagnosing an additional 1-2 patients from the remaining cases in our cohort. This data would be consistent with previously reported findings on the increased diagnostic yield of LRWGS, which is especially important in these rare and undiagnosed cases.