Researchers use long-read genome sequencing for first time in a patient 

Stanford Medicine

Stanford scientists have used a next-generation technology called long-read sequencing to diagnose a patient’s rare genetic condition that current technology failed to diagnose.

When Ricky Ramon was 7, he went for a routine checkup. The pediatrician, who lingered over his heartbeat, sent him for a chest X-ray, which revealed a benign tumor in the top-left chamber of his heart. For Ramon, it was the beginning of a long series of medical appointments, procedures and surgeries that would span nearly two decades. The trouble was, doctors couldn’t diagnose his condition.

When Ramon was 18, doctors thought his symptoms were suggestive of Carney complex, a genetic condition caused by mutations in a gene called PRKAR1A. However, evaluation of Ramon’s DNA revealed no disease-causing variations in this gene.

Now, eight years later, researchers at the Stanford University School of Medicine have used a next-generation technology — long-read sequencing — to secure a diagnosis for Ramon. It’s the first time long-read, whole-genome sequencing has been used in a clinical setting, the researchers report in a paper published online June 22 in Genetics in Medicine.

Current sequencing technologies cut DNA into “words” that are about 100 base-pairs, or letters, long, according to the study’s senior author, Euan Ashley, DPhil, FRCP, professor of cardiovascular medicine, of genetics and of biomedical data science. Long-read sequencing, by comparison, cuts DNA into words that are thousands of letters long. The type of long-read sequencing developed by the research team’s collaborators at the company can continuously spool long threads of DNA for letter-by-letter analysis, limiting the number of cuts needed.

“This is exciting,” said Ashley, “because instead of having 100-base-pair ‘words,’ you now have 7,000- to 8,000-letter words.”

Thanks to technological advances and increased efficiency, the cost of long-read sequencing has been falling dramatically. Ashley estimated the current cost of the sequencing used for this study at between $5,000 and $6,000 per genome.

Though the cost of short-read sequencing is now below $1,000, according to Ashley, parts of the genome are not accessible when cutting DNA into small fragments. Throughout the genome, series of repeated letters, such as GGCGGCGGC, can stretch for hundreds of base pairs. With only 100-letter words, it is impossible to know how long these stretches are, and the length can critically determine someone’s predisposition to disease.

Additionally, some portions of the human genome are redundant, meaning there are multiple places a 100-base pair segment could potentially fit in, said Ashley. This makes it impossible to know where to place those segments when reassembling the genome. With longer words, that happens much less often.

Given these issues, 5 percent of the genome cannot be uniquely mapped, the researchers wrote. And any deletions or insertions longer than about 50 letters are too long to detect.For patients with undiagnosed conditions, short-read sequencing can help doctors provide a diagnosis in about one-third of cases, said Ashley. But Ramon’s case was not one of those.

The technique initially used to analyze Ramon’s genes failed to identify a mutation in the gene responsible for Carney complex, though Ashley said co-author Tam Sneddon, DPhil, a clinical data scientist at Stanford Health Care who browsed through the database of Ramon’s sequenced genome by hand, did notice something looked wrong. Ultimately, the long-read sequencing of Ramon’s genome identified a deletion of about 2,200 base-pairs and confirmed that a diagnosis of Carney complex was indeed correct.

This work is an example of Stanford Medicine’s focus on precision health, the goal of which is to anticipate and prevent disease in the healthy and precisely diagnose and treat disease in the ill.