Tweet
Rare diseases affect an estimated 3.5–5.9% of the global population1. Patients dealing with these conditions not only face a direct impact on their health but also frequently encounter challenges in identifying the genetic origins of their disorder. This often leaves them with limited treatment options. However, the field of rare disease diagnosis has seen significant advances with the introduction of next-generation sequencing (NGS). Essential techniques like whole-exome sequencing (WES), whole-genome sequencing (WGS), and RNA sequencing (RNA-seq) are now routinely used for diagnosing and understanding these complex conditions. These methodologies offer a more thorough analysis of genetic variations and improve the identification and management of rare diseases. To better understand the current state of the field, we spoke with two leading experts about their groups’ contributions to rare disease diagnostics.
Expanding Diagnostic Testing
CENTOGENE is an organization committed to the diagnosis of rare diseases with a mission to facilitate effective treatment. Much of their diagnostic work involves the use of next-generation sequencing with the application of key techniques like WGS, WES, RNA-seq, and targeted sequencing. However, Prof. Peter Bauer, Chief Medical and Genomic Officer for CENTOGENE, explained that their approach focuses on the utilization of multiomics. This includes the collective analysis of different types of biological data, such as genomics, proteomics, metabolomics, and more. Bauer noted that this comprehensive approach can construct a more complete picture of a patient's condition, enhancing the accuracy of diagnosis.
While their focus is primarily on rare and neurodegenerative diseases, Bauer emphasized that due to the extensiveness of their system, it can be applied to many additional disorders that have a genetic background. They have also recently enhanced their multiomic offerings that enable filter card technology to be used for the collection and mailing of dried blood spots for subsequent extraction of DNA, RNA, enzymes, and metabolic biomarkers. Bauer believes this commitment to continuous development ensures that the company stays at the cutting edge of diagnostics. “It's not just a technology we offer,” stated Bauer. “It's our ambition to make a difference for these patients.”
The Importance of Diversity
A critical component of accurate diagnostics and representative biological research is the need for diverse datasets. One recurring theme in biomedical research has been the lack of diversity and the overrepresentation of individuals with European ancestry2. Fortunately, several initiatives in the field like the All of Us Research Program are facilitating the collection of diverse health databases for more equitable care and diagnoses.
Bauer highlighted the importance of diversity by detailing how CENTOGENE’s diagnostic capabilities are boosted by their extensive CENTOGENE Biodatabank, which contains over 800,000 cases from a wide range of ethnic backgrounds. According to Bauer, the importance isn't in the numbers but in the representation of over 120 countries, a factor that is critical for the robust interpretation and analysis of genetic variants due to the global diversity it brings. He noted that this extensive dataset allows for more precise diagnostics, particularly for individuals from underrepresented groups.
Using AI for Improved Analysis
Artificial intelligence (AI) has become a significant component of many healthcare and rare disease diagnosis workflows3. Its application can assist with difficult tasks including the analysis of larger datasets and the identification of important patterns, especially in the complex field of genomics. Mark Kiel, Chief Scientific Officer and Co-Founder of Genomenon, characterized his group’s use of AI technology by describing them as an “AI-driven genomics company” and was quick to add the value of his team’s expertise in ensuring the accuracy and clinical applicability of the data produced by their AI applications. Genomenon specializes in compiling and interpreting extensive amounts of data, resulting in a comprehensive index of genomic associations and curated genomic evidence. This index is instrumental in diagnosing diseases across thousands of clinical laboratories and facilitating drug development in dozens of pharmaceutical pipelines.
The primary data sources for this index come from clinical and scientific literature, which includes full text, figures, tables, supplemental content, and more. Kiel pointed out that by including literature from past decades and conducting weekly updates, they ensure their materials are consistently up-to-date. He went on to explain the group’s core purpose is the continual compilation and refinement of their genomic associations, which are used to link diseases, genes, variants, patient treatments, and phenotype changes. Genomenon also employs a large team of curators who focus on incorporating their AI capability and ensuring maximal sensitivity, accuracy, and specificity of their data. “We're on a mission to curate the entire genome,” Kiel stated. To this end, the company recently announced its first milestone in this journey, curating the clinical exome at the gene level.
With this wealth of information, the company designed Mastermind, a genomic intelligence platform that allows clinical users to search for evidence supporting the role of genetic variants in disease causation. Kiel noted that what makes Mastermind unique is its ability to identify a wide range of genetic variants with high sensitivity, its comprehensive and AI-driven approach to analyzing these variants, the depth and breadth of its reference database, and the sophisticated manner in which it organizes this information.
The data collected and analyzed by Genomenon is typically utilized in the clinical market and pharmaceutical development. On the clinical side, Mastermind streamlines clinical workflows and genetic testing, especially for genetic diseases and cancer, by organizing evidence around genetic variants for precise diagnosis and treatment. For pharmaceutical companies, Genomenon's data advances drug development by identifying genetic disease associations and aiding in trial design for genetic conditions.
Challenges in Diagnosis
The primary challenge in diagnosing rare diseases lies in their very name; their rarity makes determining the underlying causes a complex task. Bauer explained that to overcome this, his group works quickly to investigate the data using their own AI tools and state-of-the-art bioinformatics to find patterns. He noted that CENTOGENE has invested significant resources to analyze their data and systematically re-analyze anything that needs additional attention.
Kiel shared a similar sentiment about the challenges of diagnosing rare diseases by explaining how difficult the task is for physicians due to the large number of diseases to diagnose, particularly rare ones with unique genetic idiosyncrasies. The sheer volume of data from sequencing tests, which can now identify hundreds of diseases, complicates diagnosis and requires expert interpretation. “The interpretation of that data is still a bottleneck,” said Kiel, stressing that this is one of the reasons that he founded Genomenon. He noted that this situation is where their Mastermind software provides the “heavy lifting” of dissecting the vast amounts of data, thereby reducing the time needed for physicians and variant scientists to analyze the data and make informed clinical decisions.
Emerging Trends and Future Directions in Diagnostics
The technologies used for disease diagnosis have come a long way since the first sequencer was FDA-approved for diagnostic applications a decade ago4. Since then, several improvements and new technologies have increased the approaches used to study and identify diseases. In particular, Bauer discussed two emerging trends: long-read sequencing and advancements in multiomics. He noted that while long-read sequencing offers technical advantages, its high costs have limited its diagnostic use. However, he believes a hybrid model combining long-read and short-read sequencing could soon become viable for certain complex genetic conditions. Additionally, he shared that developments in multiomics promise enhanced resolution in small molecule profiling and quantification.
Another notable trend is the rise of collaborative projects such as the Newborn Genomes Programme, which focuses on the immediate sequencing and analysis of newborns' genomes to detect rare genetic conditions. Similarly, Genomenon has partnered with groups like Rady Children’s Institute for their BeginNGS™ program. This project involves screening newborns for genetic variants to support doctors in treatment decisions. Genomenon is working to provide vital evidence for accurate disease diagnosis in the sequencing project, ensuring data specificity and preventing false diagnoses.
The increase in technology integrations, initiatives on diversity, improvements in technologies, and the application of AI for better insights all point toward a brighter future for rare disease diagnostics. Particularly, Kiel envisions a future where AI, combined with human curation, enhances genomics research by making accurate predictions on novel variants. He believes this will drive empirical studies and allow for the integration of findings into clinical applications, creating a virtuous cycle of innovation and discovery. “We can do a lot of things at scale that we couldn't have even dreamed of before,” Kiel stated. “The scale of computation, experimentation, databasing, and dissemination of that information are all game-changing developments in a big data space like genomics.”
References
- Nguengang Wakap, S., Lambert, D. M., Olry, A., Rodwell, C., Gueydan, C., Lanneau, V., ... & Rath, A. (2020). Estimating cumulative point prevalence of rare diseases: analysis of the Orphanet database. European Journal of Human Genetics, 28(2), 165-173.
- Hindorff, L. A., Bonham, V. L., & Ohno-Machado, L. (2018). Enhancing diversity to reduce health information disparities and build an evidence base for genomic medicine. Personalized medicine, 15(5), 403-412.
- Wojtara, M., Rana, E., Rahman, T., Khanna, P., & Singh, H. (2023). Artificial intelligence in rare disease diagnosis and treatment. Clinical and translational science, 16(11), 2106–2111. https://doi.org/10.1111/cts.13619
- Collins, F. S., & Hamburg, M. A. (2013). First FDA authorization for next-generation sequencer. New England Journal of Medicine, 369(25), 2369-2371.