One of the major challenges of rare disease genomics is the availability of patient samples. The disorders being studied are, by definition, rare in the human population, and the variants that cause them (usually) rarer still. When it comes to identifying and establishing truly novel disease genes — genes that have not yet been associated with disease in humans — the burden of proof is fairly high. One strategy for providing such proof is to show that the same variant, or variants in the same gene, segregate with disease in multiple unrelated patients.
Of course, you have to find those patients first.
This should be increasingly possible in an era of widespread exome and genome sequencing. Especially among researchers who are willing to collaborate and share data.
Gene and Variant Matching Services
A number of services have emerged that aim to connect patients, families, and researchers interested in the same candidate gene in similar phenotypes.
The Matchmaker Exchange (above) acts as a centralized portal to query across GeneMatcher, DECIPHER, myGene2, and other such portals. Also, research groups such as the Centers for Mendelian Disease Genomics maintain up-to-date lists of the phenotypes and genes they’re actively pursuing.
I was surprised to learn, while attending the Family Genomic Studies Workshop at OSU this week, that many researchers working on rare diseases weren’t aware of these resources. That’s why I’m highlighting them here, briefly. What I’d really like to discuss today, however, is the power of old-fashioned communication and collaboration.
Shoe leather, in other words.
Making the Most of ASHG
The American Society of Human Genetics annual meeting takes place next month in San Diego. The abstracts are now online. Most attendees probably won’t bother reading them in advance, and I think that’s a huge missed opportunity. Here’s a story to illustrate why. I moved to Nationwide Children’s Hospital about two years ago, and began working with the already-in-progress Rare Disease Genomics Project. This was an effort to uncover the genetic basis of disease in children with rare, undiagnosed disorders who had exhausted all clinical testing.
At the time, the project had already enrolled more than a dozen cases, and had some promising leads. One of them was a particularly memorable (and sad) pediatric case. The proband was a boy who was born with arthrogryposis (contractures of the joints) and a striking atrophy of skeletal muscle. He had a host of medical issues. Whole-genome sequencing had uncovered a de novo mutation in BICD2, a gene associated with spinal muscular atrophy of the lower extremities.
The BICD2 Mutation Story
But there were two problems. First, all of the reported disease-causing mutations were missense changes (in blue, below), but our patient had an inframe deletion (in red, below) that removed a single amino acid.
Second, most BICD2 mutation carriers described in the literature had milder disease (than our patient), typically restricted to the lower extremities. We couldn’t reach a consensus, even internally, about whether or not we’d found the answer.
Fast forward to about a year ago, when the abstracts for ASHG 2017 came online. ASHG doesn’t just make them browsable, but searchable. So I spent an afternoon searching for phenotypes and genes that came to our attention through the rare diseases project, to learn if anyone else was working on them. There was a poster describing a muscular atrophy patient who carried a BICD2 mutation. In fact, it was the exact same inframe deletion as we’d found in our patient. Thanks to the wonderful ASHG app, I bookmarked the poster presentation so I wouldn’t forget.
At ASHG 2017, I found the poster during its assigned session and introduced myself to the lead author. She was an OB/GYN doing a fellowship in genetics. Basically, I wanted to first find out if we had the same patient, because sometimes individuals with rare disorders end up in multiple genetic studies (see ALS). But no, her patient was female and a few years older than ours. We exchanged contact information, and she promised to put me in touch with the PI.
Our patients, as it turned out, had a number of features in common. It was enough to convince both groups that we had a pathogenic mutation. And over the ensuing months, other publications on BICD2 — including one report of a different inframe mutation — had expanded the set of associated phenotypes. Our collaborative paper, now online at Molecular Case Studies, supports the expansion of clinical features among BICD2 mutation carriers to include cerebral atrophy, seizures, dysmorphic facial features, and profound muscular atrophy.
Sadly, as we were writing up the study, we learned that our patient had passed away. Our finding would not have presented this, but it still felt like a gut-punch. Even so, we had found an answer and helped add to the literature about BICD2 mutation carriers. I like to think that it illustrates the power of both stubbornness and collaboration when working on rare disease cases.