By Leo Barolo
Rare diseases affect nearly 1 in 12 Americans, and over a third of infant deaths are attributed to rare genetic diseases. A major challenge is diagnosing these diseases and identifying therapeutic treatments. Now, the emergence of rapid genome sequencing and specialized undiagnosed disease programs across the country is helping to diagnose patients based on their DNA sequence.
One of these programs is the UW-Undiagnosed Disease Program (UW-UDP), here at UW-Madison. The UW-UDP clinic is led by director Dr. Stephen Meyn and associate director Dr. Bryn Webb, both medical geneticists in the Center for Human Genomics and Precision Medicine.
The program offers an exciting realization of the Wisconsin Idea to serve Wisconsin patients while also leading the way in developing new technologies to improve genetic diagnosis. The UW-UDP clinic provides an opportunity for patients and families to enroll in a research study where cutting-edge technologies, such as long-read whole genome sequencing, facilitate diagnosis. April Hall, Assistant Professor of Pediatrics and the program’s genetic counselor, notes: “The UW-UDP bridges the clinical and research realms, bringing advanced technologies and the excitement of precision medicine directly to the patients at UW. Our goal is to enhance their care and well-being while also paving the way for broader advancements in precision medicine.”
Patients referred to the UW-UDP have typically already undergone extensive phenotyping and clinical genetic testing, including exome sequencing. With this technology, a diagnosis is made in ~30 to 35% of patients. The 65 to 70% who remain undiagnosed are offered the opportunity to enter a research study for complete genomic analyses beyond what is available with routine clinical care. For these patients, advanced genetic and genomic technologies are developed, applied, and refined to help identify the cause of their disease. To date, the program’s use of advanced genomic methods produced a 5 to 10% increase in diagnosis. This rate is certain to increase with new research and advancements in the interpretation of genetic variants.
New technologies advance UW-UDP diagnostics
Although current practices in genome sequencing can identify the genes responsible for causing disease, many cases remain undiagnosed. Standardly used technologies like exome sequencing may miss some causal genetic mutations, as those tests focus on the protein-coding portion of the genome that scientists know the most about. However, knowledge of how the genome encodes information, especially noncoding sequences, remains incomplete.
The UW-UDP is one of the first undiagnosed disease programs to use Oxford Nanopore Technologies (ONT) long-read genome sequencing, available in-house, as its first-tier test. Long-read sequencing differs from standard short-read sequencing by producing long stretches of DNA sequence that enable better identification of specific genetic changes. These include chromosomal rearrangements, changes influenced by repetitive sequences, or sequences from harder-to-examine areas. Additionally, ONT sequencing can identify sites of DNA methylation, chemical tags on DNA that can alter its function.
The UW-UDP research arm also incorporates other genomic modalities, including transcriptome (RNA) sequencing and other techniques depending on the case and established collaborations. This information is incorporated into the diagnostic pipeline on a case-by-case basis.
Once the patient’s whole genome is sequenced, the UW team analyzes the sequence to identify causal mutations. The patient’s genome is compared to those of their immediate family (parents and siblings) who may or may not have the disease. This can identify sequences found only in family members with the disease phenotypes. Then, after discounting genetic sequences common in the general population and thus unlikely to cause a rare disease, the effects of the remaining genetic differences are predicted. This is done using prediction algorithms, knowledge from animal models, and literature curation. Each of these variants is evaluated to determine their relevance to the patient’s symptoms.
However, the success of this approach depends on current knowledge of disease-causing sequences. A major limitation is that most disease-causing sequences remain unknown, such that there is simply no known information about their impact.
“We can detect most variants present in a human’s genome, but our ability to interpret variation is difficult. We have thousands to millions of variants that make each of us unique. We have to somehow filter through all of those variants, and the puzzle may be quite complicated! Given that this technology is new, the genetics community is still in a learning stage to see what variants are disease-causing versus benign,” continues Hall.
The limitation in interpreting genome sequence differences is driving new research on campus, including in the Center for Genomic Science Innovation (CGSI). One program led by CGSI faculty member Mark Craven is part of the “Impact of Genomic Variation on Function (IGVF)” consortium at the NIH. Craven and colleagues are one of 26 teams in the country developing new methods to understand human genetic variation.
Another collaboration of 7 faculty in CGSI is the “Genomic Variant Interpretation” project. This collaboration is integrating computational genomic approaches to both identify new disease-causing sequence differences and better understand diseases. Algorithms being developed can incorporate other information into genome analysis, including DNA methylation and RNA abundance information.
“Developing new tools to interpret genetic variation is an exciting opportunity,” says CGSI Director Audrey Gasch, who participates in both programs. “We have some exceptional strengths in computational genomics at UW-Madison that are making important advances in this area.”
Hall underscores the importance of these developments: “Computational genomics and interpretation of human sequence data are critical for diagnosis. We gather all available information to examine different variants, which can lead us to find novel causes of human disease,” she shares.
Enrolled patients consent to sharing their data with other researchers and clinicians for additional research studies. With consent, the data sequenced in the program is also uploaded to NIH databases to foster the continued expansion of databases of known disease-causing variants. Thus, individual patients can learn about their disease while contributing to the development of new approaches that can help others.
An important outcome of this data sharing is connecting physicians and researchers who have identified genetic variants in the same genes to share information they have learned. One such program is GeneMatcher, a website that matches clinicians with relevant researchers. This exchange of knowledge may lead to the identification of novel syndromes while connecting the undiagnosed disease community on a global scale.
Collaborative opportunities
The UW-UDP is always looking for new collaborations to analyze and interpret genetic information. Their vast dataset of patients with deep phenotyping information and long-read genome sequencing data, all with consent for broad data sharing, provides the template for impactful collaboration and multiple avenues of research. UW-UDP researchers are also interested in collaborating with scientists working with animal or cellular models who may perform functional studies on identified variants of interest.
Hall is very excited about the prospect of new collaborations, and interested researchers are encouraged to contact her at alhall2@wisc.edu. “Even at the lunch at the CGSI retreat, I had a lot of great conversations with attendees,” she says. “We have a lot of data and resources at UW with a lot of potential.”
“As expected, despite our best efforts, many patients still remain undiagnosed. So (we are always looking for) better ways to reanalyze their data until a molecular diagnosis is identified. This includes looking for improvements in our diagnostic strategies, whether that’s new, different types of testing or different ways of analyzing the genetic information we have available,” Hall concludes.