No topic has been more popular at Mendelspod than that of genomic medicine. This is partly an editorial decision. But it also comes to us from every direction. And it is very exciting to hear stories of new knowledge about human biology being translated into precision healthcare. We have featured the major players in the NGS tools industry and some of the newcomers working on “next next gen”. We’ve featured data analytics companies and some of the new genomic interpretation and reporting groups who are setting up shop. We have interviewed professors talking about their latest discoveries linking biomarkers to disease.
It’s easy when listening to these guests to get caught up in a personalized medicine wonderland.
One guest has stood out over the past two years. Cliff Reid, CEO of Complete Genomics, has a powerful vision of personalized medicine that is undeniable. But he also commands an exceptional grasp of “on the ground” realities. The implementation of research into the clinic is not always as straight forward as we hear.
Cliff was a cofounder of the company and started with the vision to sequence a million genomes. This meant more than research. The business model for the company depended upon doing clinical genomes. They had the fastest human genome sequencing technology, accuracy as good as any other platform, and a steadily dropping price. Running low on cash, the company sold last year to the Chinese sequencing and informatics company, BGI. Still, Cliff’s goal is the same: to sequence a million clinical genomes. And with the resources, infrastructure, and international footprint of BGI, Cliff can see even further into the horizon of genomic medicine.
I recently heard Cliff speaking on a panel titled “The Genome Moves to the Clinic” at the Burrill Personalized Medicine conference. His answers stood out boldly and often ran counter to the prevailing winds of the conference circuit. I met with Cliff this past month for a couple chats to get a more comprehensive view of his thoughts on the topic. Together we came up with these five myths of genomic medicine:
1. Genomes will be housed in EMRs.
Cliff turned heads at the Burrill conference when he said, point blank, that the genomic data won’t be kept in electronic medical records. It has been the accepted notion that at some point one of our great software companies--IBM, Google, Oracle--will bring phenotype and genotype together, will put a person’s omics profile together with their EMR. A doctor will come in the patient room, pull up the record, and see everything about the patient in one place. In fact, Cliff says, doctors are scared to death of this prospect. Why, I asked him.
Because of a group that we don’t talk about much in this industry, but with whom doctors are very familiar--the twelve person jury. Genomic data is powerful. But doctors don’t yet know how to use it. Let’s say the patient’s genomic profile resides in his/her medical record, Cliff reasons. Then the patient comes down with some illness or has an existing disease that hasn’t been diagnosed . This patient later gets it in his/her head that the illness could have been predicted or diagnosed based on the genomic profile in their EMR--which could be true. How responsible is the doctor? Genetic researchers don’t think about malpractice suits from day to day. Doctors do.
2. Genomic data is huge and will require an enormous, expensive storage facility.
Cliff gives a talk titled, “All of the world’s genomic data fits in my garage . . . and always will.” He expects to be booed off the stage at bioinformatics meetings. Here’s the gist of it:
The NSA is building a storage facility to house their data so big that it has to be located in the desert of Utah. But this needn’t be the case for Complete Genomics. Sequencing data is big now, Cliff says, because it’s raw data. But companies such as Complete are developing a new level of confidence in finished data--that is, the significant information gleaned from the raw data. And the raw data will not be needed. Check out these numbers:
If Cliff’s goal is to clinically sequence a million genomes, how much storage is needed? One genome generates 20 gigabytes of raw data. So one million genomes equals 20 petabytes. A data center that size can fit into Cliff’s current average size garage. (I know where you think I’m going. When genomic medicine goes mainstream, Cliff’s garage will grow accordingly too. No.)
In a few years, the finished data for a clinical genome will take up just 20 megabytes. That’s a decrease of one thousand times, from 20 gigabytes to 20 meg. Now scale up the sequencing of one million genomes to one billion genomes. That’s one thousand times increase. Which means the finished data will again be 20 petabytes. It still fits in Cliff’s garage.
And what about the need to go back to the raw data to question a conclusion? Cliff says it will be cheaper to just sequence again. The ultimate storage for raw data already exists, he says. It’s stored in molecules, not bits.
3. Genomic medicine will first take hold in the U.S.
We can’t even implement Obamacare, let alone a new practice of medicine. Research hospitals in the US are certainly the leaders. And it might seem wrong that the American taxpayer who has funded the majority of research not be the first to benefit. But it may be a smaller country in Europe, or say Asia where there is a national health service and a culture more open to risk taking.
Cliff says that in America the risk profile introduced by genomics causes a great deal of caution. This is not to say that there are cultures without ethics. But let’s take the notion of privacy, for instance. George Church has been campaigning tirelessly to get a group of people who will share their genomic data with his Personal Genome Project. Participants must sign what’s called an informed consent form. It’s been a painstakingly slow process.
Now consider a country like Denmark where there is a national health system and where all the citizens of the country have an electronic medical record. Just with that, Denmark sits better poised to implement genomic medicine. Add to that a Danish cultural phenomenon. All of these medical records are open to the public. This is a country where the population shares their tax filings!
Or, let’s go to rural China, where Cliff has spent time recently. Rural China? I ask. Surely the latest in precision medicine will be introduced in the huge metropolises we see covered in smog. No, Cliff says. There is an opportunity in rural China. There is no medical infrastructure there yet. Think about India when cell phones were introduced. The Indians had never pulled copper cable. They went directly to cell phone towers. The same might be true for the introduction of a new way of medicine into great portions of China that bypasses sluggish bureaucracies and inefficient infrastructure.
4. Medicine is first delivered by doctors.
This is a myth often laid bare by Eric Topol, and we see more and more questioning it. Still, Cliff insists that it’s not that well understood.
Genomic medicine will be incorporated into practice and will ultimately be implemented by doctors, yes. But not at first. There are passionate patients who have become experts at their own genomic status or that of someone in their family. Parents of children with rare diseases are acquiring the equivalent of a PhD in genomics, determined to find a diagnosis and treatment.
Jay Lake is a sci-fi writer struggling to survive colon cancer. Thousands of fans read his daily blogs which for the last few years have been mostly about his cancer. Since one of his fans told him of genome sequencing, Jay and his father, a former U.S. ambassador, have found their own way to get Jay’s genome sequenced and interpreted. They have pushed clinics and physicians to be up on the latest in genomic research. In our interview with Jay, he said genomic medicine is not clinical yet, it’s individual. Jay and others like him are the pioneers of genomic medicine. They are patients delivering medicine to doctors.
5. A revolution in genomics will change the face of healthcare.
Genomics is not the big revolution. The revolution with the big “R” is in social media, Cliff says. Even though this is the age of science, meaningful discovery still moves at a slow pace. We jumped forward with GWAS, for example, then back again. We had Herceptin and Gleevac a decade ago. As predicted by Moore’s Law, the technology moves forward fast, yet making sense of the data is slow. According to Cliff, if there’s any revolution that will change medicine, it’s not the science of biology as much as the phenomenon of social media. It is changing the way humans interact. It is changing our culture and turning existing hierarchies on their head. Patients are talking with each other. They are talking directly to researchers, and as seen in the previous myth, they are developing a different relationship with their physicians.
Remember the four “P’s” of Lee Hood’s P4 Medicine, Cliff insists. Don’t underestimate the “P” for ‘participatory.’ We haven’t yet nearly hit the tipping point on this one.
A couple runners up to this list were the issues of regulation and incidental findings. Here, Cliff’s thinking is not so revolutionary. Regarding regulation, he says it’s inevitable. This sentiment was backed up today by a letter sent by the FDA to popular DTC company just down the way from Cliff’s office, 23andMe. And on the topic of incidental findings, Cliff says that “it is beginning to settle down” with the recent recommendations of the ACMG.
Cliff acknowledges that DNA sequencing is only part of personalized medicine. Proteomics will be much more challenging to deal with. Other ‘omics’ will become the limiting factor. Still, genomics is charting the way. Cliff is part of team laying down the first railroad tracks of a new approach to medicine. And we’re not always sure just which way the tracks will run.