Skip to main content
Premium Trial:

Request an Annual Quote

Q&A: ClinSeq Head Talks About New Sequencing Tools Used to Study Heart Disease Genetics


Leslie Biesecker
Chief, genetic diseases research branch
National Human Genome Research Institute
Name: Leslie Biesecker

Age: 52
Title: Chief and senior investigator, genetic diseases research branch, National Human Genome Research Institute, since 2006
Experience and Education:
— Investigator, genetic diseases research branch, 1993-2006
— Lecturer in pediatrics, University of Michigan, 1990-1993
— Medical genetics fellowship, University of Michigan, 1988-1990
— Staff pediatrician, St. Louis Children’s Hospital, 1986-1988
— Pediatrics residency, University of Wisconsin, 1983-1986
— MD, University of Illinois College of Medicine, 1983
— BS in biochemistry, University of California, Riverside, 1979

At the National Human Genome Research Institute, Leslie Biesecker oversees the genetic diseases research branch, and leads ClinSeq, a large-scale medical sequencing study that focuses on heart disease.
In early 2007, the NIH started enrolling subjects in the study, which ultimately plans to sequence the genomes of 1,000 participants (see In Sequence 6/5/2007).
Earlier this month, In Sequence spoke with Biesecker to hear how the study has been progressing, and what sequencing technologies the researchers have explored so far.
When we spoke last year, you had recruited about 50 participants and had sequenced 39 genes in these that contribute to heart disease, using Sanger capillary sequencing. Can you give an update on the ClinSeq study?
Currently, we have recruited about 550 patients, so recruitment is going extremely well, and we have now sequenced 225 genes in about 375 patients, so our sequence acquisition rate is increasing very rapidly.
Our production is still based on PCR and capillary [electrophoresis sequencing]. But we are piloting a number of approaches for exon and eventually exome capture, including a collaboration with Jay Shendure of the University of Washington, as well as with Chad Nusbaum at the Broad Institute. And we are collaborating as well with Francis Collins’ group, who is piloting NimbleGen-based capture techniques. So we are trying to probe all three of these technologies as a pathway toward capture and next-gen sequencing for our production pipeline.
Do you have any early results on these capture methods yet?
The results are relatively modest in the number, and they are far from enough yet to tell us which way we should go. And I think it’s fair to say that all three of these methods are really themselves in a research developmental phase still. As far as nuts and bolts, making it work in a production facility and scientifically effective in a cost-effective manner, we are not ready to jump yet. We are still working on them, and exploring our options, and seeing which way is the best way to go.
What has been the greatest challenge of the study so far in terms of generating, archiving, accessing and sharing the data, detecting and assessing variants, and communicating the results?
‘Yes’ is the answer to your question. All of those are challenges, absolutely. The current production pipeline that the NIH intramural sequencing center has is sufficiently robust, [so] that our ability to generate data through that pipeline is far above the limitations of all those other facets that you listed. So we can generate data faster than it can be distributed efficiently, comprehensively analyzed, interpreted, and returned to participants.
And it was our goal from the outset to put ourselves in exactly that situation, to face the reality of high-throughput genomics in clinical research, in order to force us to scale all these other processes that are necessary to do this kind of work. We are in that situation, and that’s a good thing, because that will spur us to increase our efficiency. We are currently basically skimming — or some people would call it cherry-picking — the data for the most obvious severe and clearly pathologic variants that are in the dataset. And there are a number of those now: 6 of our 350 probands have been found to have rare high-penetrance Mendelian disorders that mimic the common phenotype of hypercholesterolemia and atherosclerosis. So they are essentially serving as positive controls, [and] it demonstrates that the whole pipeline works.
We are controlling the problem we have created by setting extremely high thresholds, if you will, for the kinds of mutations we are searching for. And our goal is to back off on the stringency of those thresholds and deal with more. That, then, gets us into another interesting phase of this, which is that then one is in the realm of detecting variants that are plausibly associated with the phenotype, but not certainly. And, as well, that gets one into clinical discovery, which is the other exciting thing that we hope to be doing with this project. We want to use this clinical genomics pipeline as a hypothesis-generating tool to help us discover new phenotypes that are admixed in with these common phenotypes that are out there.
I think the phrase that’s been invented to describe this is ‘the dark matter of genetics.’ This notion is that these very low-frequency variants are out there. There are many individually very rare variants that are contributing to the heritability of these traits that we don’t seem to be seeing with the GWAS common variant approach. And we have to start to find those, and computationally and biomedically, that’s a challenge, because that has never been done before. So we have to develop the pipelines, and we have three or four projects relating to various common phenotypes, like lipid levels, and glucose tolerance, and body mass, that we are looking at as potential avenues to explore to make those kinds of discoveries.
How are you validating your results?
We have designed the study with parallel pipelines. One is a production and discovery pipeline, and then [there is] a parallel clinical testing pipeline. As soon as the blood tubes are collected, those pathways split, and there is a separate clinical pipeline where a blood sample goes for DNA isolation that never has anything in common with the production pipeline. That goes to a CLIA-certified clinical testing laboratory setting, and it awaits a discovery from the discovery pipeline, and when we have a finding, we then go to the clinic coordinator, who contacts the subject to learn if the subject — we describe vaguely what the finding is — would want to know of such a result. Then if they do, we tell them we will work further and contact them soon. Then that triggers the clinical testing pipeline to start working on that variant from the clinical DNA from that subject. Then that finding is replicated in a clinical testing lab, a clinical test report is generated, that is sent to the clinical coordinator, and then I and the clinic coordinator counselor schedule an appointment for the subject to come in, and we disclose the result and provide genetic counseling, medical counseling, and help them initiate the therapy that they need.
And that pipeline I just described to you is for the finding of variants in genes that are already known today to cause the phenotype. So we have a handful in LDLR, the low-density lipoprotein receptor. Not a novel finding, but it demonstrates that the pipeline can find these things, which is great.
For the next category of disorders, where we will hopefully discover new relationships of genotypes to phenotypes, we have an additional step that we have to go through, which is a mutation advisory panel, where we will perform whatever clinical research we think is necessary to make our case that we think this set of variants causes the phenotype in our research subjects. We have to present those data to the advisory panel, which is a group of scientists and physicians who are not part of the study, convince them that our findings are sufficient to warrant reporting back to the subjects, [and] if they approve it, then that triggers that pipeline to start. So that provides an extra measure of protection for the human subject from false-positive research findings.
Have you planned any replication studies?
Yes. Those will be done on an ad-hoc basis, because it will depend on exactly what the phenotype is that we are looking at, and we will go out to different groups and start collaborations and validate and replicate findings that way.
What are you planning to do over the next near or so?
Our goal is to switch to a method of capture and next-gen sequencing for production within the next year. We want to do that because we are very excited about the capacity of such a system. We have a candidate gene list, but we are impatient people, and we want to know about more data and more genes. And we think that’s what the excitement about genomics is all about, is throughput and parallelism.
Is the goal still to scale up to whole-genome sequencing? When do you see that on the horizon?
It’s really hard to answer that question. It does get down to, of course, cost. And when there is a platform that is ready to generate whole-genome sequencing on our subjects in the range of $5,000, we would migrate to that. And I find it personally hard to predict when that will occur. I don’t know exactly when that’s going to be available to me. But the subjects are already consented. The samples are ready to go to whatever platform becomes available the day after it’s available to us.
Is it true that you have seen an enthusiastic response to your study so far?
Yes, we can’t keep up with our demand. We are currently able to bring 12 patient a week through the clinical center with our medical staff, and we have waiting lists. [The subjects are] mostly from metropolitan Washington [though] I just saw a lady today from Baltimore. We thought we would have to advertise and recruit aggressively from Baltimore and even Richmond, but we don’t. We have plenty of people in metropolitan Washington who are interested. The response has been very enthusiastic.
Is ClinSeq still the only study of its kind?
[There are] studies that are occupying various points in what I call the ‘translational genomics research space,’ which is sort of a three-dimensional space: on one axis, you have a number of subjects, on a second axis, you have breadth of genome acquisition, and on the third axis, you have phenotype depth. You take studies — if you can call them that — of the personal genomes that are now out there, of which there are now four or five, and those are all the way out on the genome axis, N’s of one each as far as the subjects, and sketchy phenotypic data. Then there is [George] Church’s PGP [Personal Genome Project], which is purported to be 100,000 patients with available medical records, but I don’t think they are thinking about actually phenotyping their patients themselves. So that would be lower on the phenotype axis, but higher in the N. And there is going to be a number of studies like this, and we’ll just have to see where all of these studies make those compromises of those three key variables. And we are each going to learn different things from those studies, because they are structurally differently.
Are you planning to publish results soon?
We are hoping to publish what we are going to call a ‘marker paper,’ if you will, which basically describes how the study is constructed and then gives some preliminary results, hopefully within the next few months.

File Attachments
The Scan

UK Pilot Study Suggests Digital Pathway May Expand BRCA Testing in Breast Cancer

A randomized pilot study in the Journal of Medical Genetics points to similar outcomes for breast cancer patients receiving germline BRCA testing through fully digital or partially digital testing pathways.

Survey Sees Genetic Literacy on the Rise, Though Further Education Needed

Survey participants appear to have higher genetic familiarity, knowledge, and skills compared to 2013, though 'room for improvement' remains, an AJHG paper finds.

Study Reveals Molecular, Clinical Features in Colorectal Cancer Cases Involving Multiple Primary Tumors

Researchers compare mismatch repair, microsatellite instability, and tumor mutation burden patterns in synchronous multiple- or single primary colorectal cancers.

FarGen Phase One Sequences Exomes of Nearly 500 From Faroe Islands

The analysis in the European Journal of Human Genetics finds few rare variants and limited geographic structure among Faroese individuals.