Last month, Georgetown University Medical Center launched a new biomedical informatics center that will serve as an academic hub for individualized healthcare research efforts.
GU tapped Subha Madhavan, director of clinical research informatics at GU’s Lombardi comprehensive cancer center, to serve as the inaugural director of the Innovation Center for Biomedical Informatics (BI 9/7/2012).
The center will develop tools to help researchers and clinicians evaluate genomic data in clinical and translational research studies. The ICBI team will also conduct research studies focused on diseases other than cancer and provide broader support to the medical center.
The new ICBI is an extension of the institute’s clinical informatics group, which developed the Georgetown Database of Cancer — a repository of cancer information and tools (BI 10/29/2010). Since its launch, G-DOC has been used in internal and external collaborations at the university and has been extended to support research in disease areas outside oncology, including Alzheimer’s disease and muscular dystrophy.
So far, the center has received funds from the National Institutes of Health, the US Food and Drug Administration, and undisclosed corporations for personalized medicine, drug discovery, and pharmacogenomics projects.
BioInform spoke to Madhavan last week to learn more about ongoing ICBI projects. What follows is an edited version of the conversation.
Have you taken up the reins at the new center?
Yes. We actually were the Clinical Informatics group within Lombardi Cancer Center at Georgetown since 2008, working primarily on cancer genomics. We were officially approved to become a center this summer under the umbrella of the Georgetown University Medical Center. Our portfolio of projects has grown into non-cancer disease areas as well, so the center allows us to expand our purview and enhance collaborations within the community.
What have you been working on since you got started?
ICBI is really an academic hub for cutting-edge translational research. What we do at the center is bring together multidisciplinary groups of people who are knowledgeable in the fields of computer science, biomathematics, biostatistics, genomics, clinical data interpretation, and clinical decision making … to address biomedical research problems especially in the areas of clinical and translational research and, hopefully, to ultimately impact clinical care.
I’ve been here for four years conducting research and providing services in the area of biomedical informatics at Georgetown and this was a natural transition. Our initial success within the cancer center has allowed us to leverage this as a platform to support the medical center more broadly. We are starting to work on projects on dementia, Alzheimer’s, preterm birth, pediatric muscular dystrophy efforts, and even the technological challenges of dealing with big data, which are taking us beyond the space of clinical and translational research in cancer alone. That said, about 75 percent of the projects we work on are in the cancer field, though we are slowly starting to ramp up our work in other disease areas as well.
Was it because you wanted to look into projects outside of cancer that the university decided to expand the informatics group?
That’s exactly the reason. We wanted to support the medical center more broadly [and] leverage the developments in informatics to other disease areas. Our platforms are scalable and disease agnostic.
You’ve received some funding from a number of agencies. Can you tell me what you plan to use the money for?
We have funding from the FDA, for example, from the office of the chief scientist [under] the [Office of Regulatory Science and Innovation]. We are conducting a number of informatics-oriented research projects in collaboration with FDA investigators. One project is around pharmacogenomics in cancer and how particular genotypes can affect drug metabolism … and the other area is developing better evidence for vaccine adverse events reporting. We are trying to increase the evidence base of the data that’s in the vaccine adverse events reporting system. We are trying to enrich the data using protein and gene annotation and other molecular information — [for example,] pathway information — to really understand what adverse events are directly correlated with vaccines. This is a critical project for ICBI in that it really allows us to leverage our multi-disciplinary team to enhance regulatory decision making at the FDA through informatics research.
There are a number of other efforts. For example, we have funding from the National Cancer Institute for the Center for Cancer Systems Biology. This is a project that the Lombardi Cancer Center has had for about three years now and the informatics activities of this effort have naturally transitioned into an ICBI effort. This is a breast cancer-focused project to conduct computational as well as mathematical modeling in breast cancer to understand tamoxifen resistance. ICBI acts as the informatics core for the CCSB effort.
One more example is our Clinical Translational Science Award effort. Through our CTSA, and part of the biomedical informatics component that sits within ICBI, we are rolling out data integration [and] data collection platforms. One of the new projects that we have embarked [on] recently is to engage the patient community in clinical and translational sciences. For that, we are leveraging an iPad-based mobile application called Tonic Health and we are customizing that platform to enable collection of data when the patient visits the clinic. The idea is to collect data once and use it multiple times. We are working with Tonic Health to collect data in a much more standardized form that can be de-identified for downstream use in clinical and translational research projects.
We have some other corporate engagements that are beginning to evolve as well. These are in early stages.
Sounds like you have a lot of projects going on. Have you had to expand your team?
We have two positions open at this point. Our group is about 20. We have gradually grown because of support both from the cancer center as well as across the medical center and the external project funding that we have received over the course of the last four years. But there are areas that we would like to tap into additional talent, some of the areas being big data, for example. Next-generation sequencing is a core area of focus for us from a methodology development perspective and a data analysis perspective.
So, we are looking to recruit a faculty member that can really spearhead that area of NGS informatics and correlating NGS data [such as] variant information and structural variation data with clinical genotypic information. We have a number of projects both at the cancer center and outside that are leveraging NGS platforms. We have some expertise locally but we are really looking to recruit another faculty member. We also have a bioinformatics software developer position open to work on mobile technology for … the Tonic Health project [and] expanding that notion of making data really come alive in a very dynamic fashion to [give] physicians and patients access to this information.
How about your infrastructure? Have you had to beef up your hardware?
We have a hybrid approach to infrastructure at ICBI. We leverage Georgetown University’s information services for our G-DOC platform, [for example,] which is now a widely used platform with about 350 registered users … across six different countries. But for some of our other collaborators who require much larger data storage on projects that have 500 terabytes of data from next-generation sequencing, for example, we leverage Amazon cloud computing. That allows us to scale very quickly so our developers don’t have to depend on infrastructure availability and we do not have to plan or pay for them ahead of time. Use as we go works well for large data storage and computes.
We also have a collaboration with Oak Ridge National Laboratory and they have a very good high-throughput computing environment that they make available to our CTSA so we are looking into leveraging some of that for our high-throughput screening and drug discovery projects.
How about software?
We continue to develop G-DOC and in fact have just released version 2.0, which includes a new clinical workflow and variant search capability among other features.
We have also pulled out the backend pieces of G-DOC and created a generic data integration platform called G-CODE [that] stands for Clinical Omics Development Engine. That is a strategy that I am working on with our office of technology transfer to make that platform, at least the core modules of it, open source to the biomedical research community. The intent is to use this as a toolkit to quickly develop various disease-specific data integration portals for research use. And a longer-term goal is to make these portals useful not only to researchers but to physicians to enable better therapeutic decision making based on the large-scale analysis of molecular data.