Skip to main content
Premium Trial:

Request an Annual Quote

Q&A: SynapDx's Mark DePristo on Informatics for ASD Diagnostics; Leaving Broad Institute


Mark Depristo.JPGSynapDx said this week that Mark DePristo has joined the Lexington, Mass.-based company as vice president of informatics.

In this role, he will be responsible for designing and implementing a bioinformatics strategy that will support SynapDx's proprietary test for the early detection of autism spectrum disorders. The company is currently recruiting participants for the SynapDx Autism Gene Expression Analysis, or STORY, clinical trial, to evaluate the blood test which provides an ASD risk score based on measured differences in RNA expression. Overall, it hopes to test 660 children at 20 clinical sites including Baylor College of Medicine, Children's Hospital Boston, Children's Hospital of Philadelphia, and Rush Medical Center.

Prior to joining SynapDx, DePristo led the genome sequencing and analysis arm of the Broad Institute's medical and population genetics program. His team created the Genome Analysis Toolkit, a software package that's widely used by researchers and clinicians in both academia and industry to analyze next-generation sequencing data. The GATK has been used to analyze data from large- and small-scale studies such as the1000 Genomes project and the Cancer Genome Atlas.

Before the Broad, DePristo was a research fellow at Harvard University. He also has a BA in computer science and mathematics from Northwestern University and a PhD in biochemistry from Cambridge University.

This week, BioInform spoke with DePristo about leaving the Broad and about his new role developing informatics for diagnostic testing. What follows is an edited transcript of the interview.

Has the transition from more of an academic setting to industry been a relatively smooth one for you?

When I finished my post doc, I actually joined a biotechnology consulting firm in Boston which I was at until I went to the Broad. So I was already prepared even five years ago. In the long run, I really liked working on a collaborative team building products and I wanted to be on the business side of things. Obviously the Broad was – and remains – a very interesting hybrid model between academia and industry and so it was a good home for me. I don't feel like it's been a fundamental change in my mindset, more just that the opportunity to be part of SynapDx arose and it was just too good to pass up. I feel very comfortable here.

How long were you at the Broad?

I'd been at the Broad five years. I joined in October 2008. It basically corresponded to the earliest availability of the Illumina Solexa sequencer. I was called by the Broad to help with the 1000 Genomes Project when the first data was coming off the sequencers.

Why was this a good time to leave?

As I mentioned, I joined the Broad to help with the 1000 Genomes Project. [That's] really at the end of the analysis of the final dataset [and] the bulk of the contribution that I had been making to the project [has] already been finished. There's obviously paper writing and a couple of final meetings to go to and I'll be going and helping in that effort.

The other aspect of my role at the Broad was to lead a team that was developing tools to analyze next-generation sequencing data. My role actually fell into the germline genetics realm. I think we took the GATK from an idea with no actual instantiation and implementation to the world-leading toolkit that it is today. And in the germline world at least, it's fairly mature.

The final bit of this is … I really had the opportunity to recruit and then help grow two exceptionally good junior leaders there in Eric Banks and Mauricio Carneiro – who respectively led methods development and technology development for my group. Eric Banks [has stepped into] my previous role. So there has been an exceptionally smooth transition at the Broad.

Why was SynapDx a good fit for you? Did you have experience with neurdevelopmental disorders prior to this?

There are a couple of reasons. One is that it's very exciting to be a member of a company focused on laboratory-developed tests; you have an opportunity to see your work appear in the marketplace in a shorter time frame than you would have if you were in a pharmaceutical company. Two is we know that neuropsychiatric disorders are a huge unmet need. Being able to create a test that can help [guide] detection so that more kids who are affected by autism spectrum disorders can potentially receive earlier treatment and intervention can really radically improve the lives of many people.

It’s a company that is using NGS to do a variety of biomarker assays, so we do RNA-seq, DNA-seq … a variety of things, in an attempt to define and build up the best biomarkers and detection systems we can. That’s something that I see as a natural expansion of my role in the germline genetics side of the Broad. The other side is my little brother is actually on the ASD spectrum himself so there's a personal connection.

What will your responsibilities be as SynapDx's vice president of informatics?

We have a clinical trial underway right now called STORY that’s [rigorously] evaluating a signature that we've seen in a variety of other datasets. One responsibility I have is to shepherd through the analytics of that process. That involves everything from helping to understand how [to] do the best experiments that we can, how [to] interpret the data in the most rigorous way, [and] really define the signature so that when we unblind the study for the validation set we do as well as we can.

The other aspect is CLIA validation and CLIA lab certification. That’s a new activity for me.

Finally, there is a broadening of the platform that SynapDx is using to define biomarkers. We are really thinking of ourselves as investing in the infrastructure and knowledge of how to do RNA-seq and DNA-seq so that we can succeed on these kinds of biomarkers that we want to use for diagnostics. So there is the question of defining what this platform looks like, what are the pipelines, what's the analysis system, [and] how do you integrate that type of data together? That’s a new and very interesting thing for me.

Are there existing tools that would be sufficient for your needs?

I think in the germline genetics side, there are very good tools in that space. RNA-seq is also a fairly mature area, so there are existing tools. I think it’s more a question of … if you want to be able to do the best you can at assaying gene expression in samples, this is a matter of understanding what the state of the art is, defining what you specifically need to do with those frameworks and tools that are out there, and then being able to execute on that at scale … and then just staying on the front of that wave. What's possible to do today is going to be simpler a year from now; and there'll be things that we can do a year from now, that we can't even conceive of doing now. So making sure that we stay cutting edge and that we have the analytics and capabilities in place to do that is really a very interesting challenge.

Where we are today is good but I think we will end up innovating in many different ways analytically and technically just to keep up and to satisfy our goals as a company.

What are you doing on the hardware side?

As a company, we've gone all in on the cloud. The fact that you can ramp up quickly on the cloud from a small computational and data storage footprint into huge amounts of consumption is a very attractive thing for us as we do anticipate large amounts of data processing and computing needs as we scale up. We see cloud computing as the solution to that problem. We are very much in the business of trying to understand what the best way of utilizing the cloud is [and] what the best approach is to handling genomics data on the cloud. Do we have our own platform; do we collaborate and use other platforms; do we use open source solutions; do we work with a company?

Are there any unique challenges associated with analyzing autism spectrum data in particular?

The thing that makes autism spectrum disorders on the scientific side challenging is it's an example of a classic disease that is highly heritable; it has a significant phenotype; it’s clearly a very heterogenous disorder with multiple subcategories. It’s a complicated disorder.

When does SynapDX plan to have a test on the market?

The STORY clinical trial is underway and we're pleased with how recruitment is going [but] at this stage we really can't project when the study will be done.