Periodic Reporting for period 4 - Solve-RD (Solving the unsolved Rare Diseases)
Periodo di rendicontazione: 2022-07-01 al 2023-09-30
“Solve-RD – solving the unsolved rare diseases” is a research project funded by the European Commission for five years (2018-2022). It echoes the ambitious goals set out by the International Rare Diseases Research Consortium (IRDiRC) to deliver diagnostic tests for most rare diseases by 2020. The current diagnostic and subsequent therapeutic management of rare diseases is still highly unsatisfactory for a large proportion of rare disease patients – the unsolved rare disease cases. For these unsolved rare diseases, we are unable to explain the etiology responsible for the disease phenotype, predict the individual disease risk and/or rate of disease progression, and/or quantitate the risk of relatives to develop the same disorder.
Our main ambitions are thus
• to solve large numbers of rare diseases, for which a molecular cause is not known yet by sophisticated combined omics approaches, and
• to improve diagnostics of rare disease patients through contribution to, participation in and implementation of a “genetic knowledge web” based on shared knowledge about genes, genomic variants and phenotypes.
Solve-RD fully integrates with the newly formed European Reference Networks (ERNs) for rare diseases. Four ERNs (ERN-RND, -EURO-NMD, -ITHACA, and -GENTURIS) build the core of Solve-RD but we will reach out to patient cohorts across all 24 ERNs as well as the undiagnosed disease programs in order to achieve our aims.
Solve-RD identified 3 main challenges and will deliver 7 implementation steps to address these challenges in work packages 1-3:
Challenge 1: Accessibility of unsolved rare disease cohorts with comprehensive genetic and phenotypic data
Challenge 2: New and improved approaches for the discovery of novel molecular causes
Challenge 3: Translate discoveries to patients’ live and clinical practice
Our main ambitions are thus
• to solve large numbers of rare diseases, for which a molecular cause is not known yet by sophisticated combined omics approaches, and
• to improve diagnostics of rare disease patients through contribution to, participation in and implementation of a “genetic knowledge web” based on shared knowledge about genes, genomic variants and phenotypes.
Solve-RD fully integrates with the newly formed European Reference Networks (ERNs) for rare diseases. Four ERNs (ERN-RND, -EURO-NMD, -ITHACA, and -GENTURIS) build the core of Solve-RD but we will reach out to patient cohorts across all 24 ERNs as well as the undiagnosed disease programs in order to achieve our aims.
Solve-RD identified 3 main challenges and will deliver 7 implementation steps to address these challenges in work packages 1-3:
Challenge 1: Accessibility of unsolved rare disease cohorts with comprehensive genetic and phenotypic data
Challenge 2: New and improved approaches for the discovery of novel molecular causes
Challenge 3: Translate discoveries to patients’ live and clinical practice
From 07/2022 to 09/2023, Solve-RD continued to implement the activities addressing the three challenges.
Although the data collection and submission had already been completed in the last reporting period, a few additional unsolved rare disease (RD) cases have been submitted to the RD-Connect GPAP. In total, 21,422 RD datasets (phenotypic & exome/genome sequencing data) have been collected for the Solve-RD re-analysis. Standardised phenotypic information (HPO, ORDO & OMIM encoded) has also been collected for 2,593 cases submitted for novel omics analysis.
The development of the rare disease cases ontology (RDCO) has been completed. The phenotypic similarity-based approach for variant prioritization for unsolved RD has been applied to all unsolved datasets from data freezes 1-3 and the analysis of results is currently ongoing. Clusters of similarity can be visualized in the newly developed OrphaScape tool.
All unsolved datasets have been re-analysed by the Data Analysis Task Force working groups. Results have been prioritized and interpreted by the Data Interpretation Task Forces of each ERN. A manuscript with the results of the re-analysis of data freeze 1 and 2 data is currently in review for publication in Nature Medicine: re-analysis and ad-hoc expert review for about 10,000 undiagnosed individuals from more than 6,000 families resulted in overall diagnostic yield of 12.6%. Systematic re-analysis contributed to this by establishing a genetic diagnosis in 8.5% of the families.
Service providers have been contracted for all novel omics technologies. Due to recent technological developments, a promising new technology termed ‘optical genome mapping’ (OGM) was added to the Solve-RD novel-omics portfolio. In total, 4,476 samples have centrally been collected in Nijmegen, QC’ed and then further distributed to the respective service providers. 3,707 samples have been sequenced and raw data has been returned. Data analysis and interpretation for the novel omics is currently ongoing.
The RDMM-Europe brokerage service connecting Solve-RD partners who discovered novel RD genes with model organism scientists that have the expertise to functionally validate these genes and variants opened 10 calls for Connection Applications. 36 applications were positively evaluated of which 33 received a respective Seeding Grant funding. For the first time in Europe we have established a novel brokerage structure connecting clinicians, gene discoverer and basic researchers in a highly flexible and efficient way to quickly verify novel genes and disease mechanisms.
In a clinical utility study, it has been tested whether WGS is a better first-tier genetic diagnostic test than current standard of care for neurodevelopmental disorders. WGS revealed variants that escaped routine tests, but the overall diagnostic yield did not significantly increase. Yet, WGS provided a more uniform, n=1 workflow that did increase diagnostic efficiency.
Using the Treatabolome database, IEMbase, ClinGen, and international cancer guidelines, we identified 73 affected individuals (14.4% of the 506 individuals diagnosed in cohort 1 data freeze 1 and 2) that harboured causative variants in a potentially actionable gene.
The data flow system has continuously been adapted to the project’s needs involving GPAP, the EGA, omics service providers, the sandbox and RD3. A scientific publication has been prepared that describes the design concept and progressive maturation of the Solve-RD data infrastructure, software and standards created by WP4. This paper is now under review in GIGAScience.
Although the data collection and submission had already been completed in the last reporting period, a few additional unsolved rare disease (RD) cases have been submitted to the RD-Connect GPAP. In total, 21,422 RD datasets (phenotypic & exome/genome sequencing data) have been collected for the Solve-RD re-analysis. Standardised phenotypic information (HPO, ORDO & OMIM encoded) has also been collected for 2,593 cases submitted for novel omics analysis.
The development of the rare disease cases ontology (RDCO) has been completed. The phenotypic similarity-based approach for variant prioritization for unsolved RD has been applied to all unsolved datasets from data freezes 1-3 and the analysis of results is currently ongoing. Clusters of similarity can be visualized in the newly developed OrphaScape tool.
All unsolved datasets have been re-analysed by the Data Analysis Task Force working groups. Results have been prioritized and interpreted by the Data Interpretation Task Forces of each ERN. A manuscript with the results of the re-analysis of data freeze 1 and 2 data is currently in review for publication in Nature Medicine: re-analysis and ad-hoc expert review for about 10,000 undiagnosed individuals from more than 6,000 families resulted in overall diagnostic yield of 12.6%. Systematic re-analysis contributed to this by establishing a genetic diagnosis in 8.5% of the families.
Service providers have been contracted for all novel omics technologies. Due to recent technological developments, a promising new technology termed ‘optical genome mapping’ (OGM) was added to the Solve-RD novel-omics portfolio. In total, 4,476 samples have centrally been collected in Nijmegen, QC’ed and then further distributed to the respective service providers. 3,707 samples have been sequenced and raw data has been returned. Data analysis and interpretation for the novel omics is currently ongoing.
The RDMM-Europe brokerage service connecting Solve-RD partners who discovered novel RD genes with model organism scientists that have the expertise to functionally validate these genes and variants opened 10 calls for Connection Applications. 36 applications were positively evaluated of which 33 received a respective Seeding Grant funding. For the first time in Europe we have established a novel brokerage structure connecting clinicians, gene discoverer and basic researchers in a highly flexible and efficient way to quickly verify novel genes and disease mechanisms.
In a clinical utility study, it has been tested whether WGS is a better first-tier genetic diagnostic test than current standard of care for neurodevelopmental disorders. WGS revealed variants that escaped routine tests, but the overall diagnostic yield did not significantly increase. Yet, WGS provided a more uniform, n=1 workflow that did increase diagnostic efficiency.
Using the Treatabolome database, IEMbase, ClinGen, and international cancer guidelines, we identified 73 affected individuals (14.4% of the 506 individuals diagnosed in cohort 1 data freeze 1 and 2) that harboured causative variants in a potentially actionable gene.
The data flow system has continuously been adapted to the project’s needs involving GPAP, the EGA, omics service providers, the sandbox and RD3. A scientific publication has been prepared that describes the design concept and progressive maturation of the Solve-RD data infrastructure, software and standards created by WP4. This paper is now under review in GIGAScience.
Solve-RD addresses the biggest challenge of diagnosing patients with RD since the implementation of NGS technology. Despite extensive studies of WES in numerous RD patient cohorts, at least 50% of all patients remain unsolved. By applying comprehensive advanced bioinformatics algorithms in four major patient cohorts we anticipate to increase diagnostic yield by about 3-5%. So far, systematic cutting-edge research-driven re-analysis and ad-hoc expert review for about 10,000 undiagnosed individuals from more than 6,000 families resulted in overall diagnostic yield of 12.6%. Systematic re-analysis contributed to this by establishing a genetic diagnosis in 8.5% of the families.
The extension of DNA analysis from WES to WGS in >2,500 well characterized patients is expected to raise this yield ultimately to about 60%. WGS bioinformatic tools that have been developed and applied within the RD-Connect framework culminated in standardised analytical approaches to analyse genome data for structural variants. Also, no standardised multi-omics approach existed at the time Solve-RD started. Solve-RD has developed and applied specific strategies for the different omics approaches to analyse specific patient cohorts in a tailored manner.
The connection of sophisticated diagnostic approaches will only be successful with deep phenotyping of patient cohorts. All 4 core ERNs have defined and selected cohorts of unsolved patients with highly peculiar (ultra-rare) phenotypes for genome sequencing and analysis. The number of patient samples included in the ultra-rare cohort is currently 485, and several publications have already proven that the rationale to find genes and disease mechanisms using this concept is highly successful. Additionally, the two newly associated ERNs ERN-Epicare and ERN-RITA contributed 160 patient samples for WGS analysis. While these patients have been selected and the WGS sequences are available and analysed, the interpretation of some of the identified variants is still ongoing.
The extension of DNA analysis from WES to WGS in >2,500 well characterized patients is expected to raise this yield ultimately to about 60%. WGS bioinformatic tools that have been developed and applied within the RD-Connect framework culminated in standardised analytical approaches to analyse genome data for structural variants. Also, no standardised multi-omics approach existed at the time Solve-RD started. Solve-RD has developed and applied specific strategies for the different omics approaches to analyse specific patient cohorts in a tailored manner.
The connection of sophisticated diagnostic approaches will only be successful with deep phenotyping of patient cohorts. All 4 core ERNs have defined and selected cohorts of unsolved patients with highly peculiar (ultra-rare) phenotypes for genome sequencing and analysis. The number of patient samples included in the ultra-rare cohort is currently 485, and several publications have already proven that the rationale to find genes and disease mechanisms using this concept is highly successful. Additionally, the two newly associated ERNs ERN-Epicare and ERN-RITA contributed 160 patient samples for WGS analysis. While these patients have been selected and the WGS sequences are available and analysed, the interpretation of some of the identified variants is still ongoing.