BERLIN – Short-read whole-genome sequencing has already led to diagnoses for dozens of families with rare diseases who had remained undiagnosed after exome sequencing, and there will likely be more cases solved with further interpretation of the data, according to results presented here on Sunday at the European Human Genetics Conference.
According to Steven Laurie, a data analyst at the National Center for Genomic Analysis (CNAG) in Spain, 90 out of 836 or about 10 percent of families who were analyzed by short-read genome sequencing as part of Solve-RD have so far received a diagnosis. The goal of the pan-European rare disease research program is to identify disease-causing variants in tens of thousands of undiagnosed families.
Solve-RD kicked off in 2018 with €15.4 million ($16.7 million) in funding from the European Union's Horizon 2020 program that recently expired after more than five years. The goal was to use the pooled expertise of several European reference networks (ERNs) and clinical as well as exome sequencing data for more than 19,000 unsolved rare disease cases to come up with additional diagnoses.
For a subset of these, the consortium also generated new datasets, including short-read and long-read genomes as well as RNA sequencing, epigenomic, metabolomic, and proteomic data.
In his presentation, Laurie focused on results from short-read genome sequencing of 836 families recruited for the project by seven ERNs. The sequencing data was generated by China's BGI on the MGI DNBseq-G400 platform, he said.
A data analysis task force for the project used a variety of tools to generate variant shortlists that included de novo and splice-altering SNVs, indels, mitochondrial variants, copy number variants, structural variants, short tandem repeat expansions, and mobile element insertions. Those variants, which mostly fell into candidate genes provided by the ERNs, were then interpreted by a data interpretation task force that included clinicians and geneticists from the ERNs.
One aspect that helped to "make real progress" solving patient cases, Laurie said, was getting the experts in the same room in a so-called "solvathon" or collaborative data interpretation workshop, which lasted several days and also helped build relationships.
As an example, he pointed out a cohort of 58 patients with childhood-onset neuromuscular disease that was submitted for analysis by Sant Joan de Déu Barcelona Hospital. Of these, 23 cases, or 40 percent, could be solved by short-read genome sequencing, which he said required confirmation by RNA sequencing in four cases.
Not all the short-read genome data has been fully interpreted yet, he said, especially with regard to CNVs, SVs, and noncoding SNVs. "We expect many more diagnoses" from the current data, he added.
Other projects within Solve-RD are still in the works or have recently been published. For example, researchers at the University of Tübingen in Germany and their colleagues published a study in the European Journal of Human Genetics last month showing that improved structural variant calling from exome sequencing data resulted in a small additional diagnostic yield of 0.4 percent of more than 6,000 unsolved cases. Another group, led by researchers at Radboud University Medical Center in the Netherlands, published a preprint in MedRxiv last month about finding likely disease-causing variants in 13 percent and candidate disease-causing structural variants in another 4 percent of 114 unsolved families using Pacific Biosciences HiFi long-read genome sequencing.
According to Laurie, working together has turned out to be the most valuable aspect of the Solve-RD program. "The mantra for rare disease research is collaboration, collaboration, collaboration," he said.