MIDAS Steering Committee Meeting Minutes

May 20, 2005
Natcher Conference Center
Bethesda, Maryland

Presentations and documents associated with this meeting are available at https://www.epimodels.org/midas/about.do Link to external Web site

Action Items

Ensure that Tom Marr has background materials for Big Iron group
Identify the next chairman of the steering committee
Draw up a document defining the terms and length of service of the members
Set up regular Principal Investigator meetings
Revise and send out for comment the Associated Projects plan

Establish regular times for PI meetings

History of MIDAS and Role of Steering Committee

Irene Eckstrand and Simon Levin reviewed the program's history including its origin in consultations organized by the Fogarty International Center (2001-2002), the first program announcement in 2003, and funding of 4 projects in May 2004. They noted that a new program announcement was released in November 2004 and that the new applications have been received at NIH and will be reviewed this summer. The earliest funding date is December 2005.

The Steering Committee first met in May 2004 and in the past year has developed MIDAS policies and established several subcommittees in addition to providing direction and focus to the Network.

Discussion of MIDAS's direction brought out several important points:
  • The PIs feel different degrees of alignment with current directions. For example, one project is pursuing what it proposed; another has had to change direction substantially.
  • Considerable collaboration and teamwork have developed within the Network. Each group has brought unique resources and talents to the MIDAS program.
  • Involvement in policy issues is one of the attractions of MIDAS, and the work should be made available and useful to policy makers. On the other hand, the committee recognized that MIDAS is a research project and that it is not productive to continually pursue "just in time" policy analyses. Involving policy makers, the end users of MIDAS research, is important and requires trust and dialog.
  • It has been frustrating to discern who will use the information and knowledge MIDAS produces. The health policy infrastructure is somewhat diffuse, and the Network is still trying to establish good contacts and communications in the appropriate places.
  • Both research and policy making need to account for changes in population structure and should include environmental factors.
MIDAS continues to need involvement and guidance from the steering committee; however, the intensity of interactions could well diminish in the coming year. The committee is very important in expanding the knowledge base of the research groups, engaging the user communities, providing direction and priorities, and supporting outside consultations.

Within the next year, a new chairman for the committee must be identified. The committee recommended establishing 3-year terms for steering committee members.

Trip Reports

Stephen Eubank and Don Burke reported on their trips to Japan and Thailand respectively. Both trips are leading to productive collaborations, potentially leading to funding from Fogarty International Center. The trips are summarized in more detail in the March 2005 MIDAS Network Meeting Minutes.

Bruce Weir and Skip Garner visited RTI with the intent of interacting with staff and planning for increased capacity. With respect to providing modelers with unimpeded access to resources, Bruce and Skip were please to see a demonstration of spillover to the National Center for Supercomputing Applications (NCSA). They also felt there was good progress in compiling results stably and securely.

Progress Reports

Informatics Group
The informatics group has made considerable progress in developing computational resources, including limited production capacity at RTI, transparent access to external supercomputers, collection of usage statistics, a data warehouse, and specialized tools. The current RTI cluster configuration includes two clusters with redundancy for both computing power and spillover management.

The vision is that RTI will become the entry point to large supercomputing facilities, and to that end, RTI is exploring several kinds of contractual arrangements with outside providers. The agreement with NCSA is farthest along, though TACC and IBM-On-Demand are also being consulted.

RTI has adopted a Use Case approach, meaning that software and architecture are described from the end users' standpoint. RTI is continuing to develop and refine its vision and system requirements, including a model repository and query system and a data management system. In addition, RTI has taken the lead in developing videoconferencing for the research groups.

Research Projects
The research groups have used somewhat different approaches to model outbreaks and interventions for avian influenza in SE Asia. It was good news that Ira Longini and Neil Ferguson have submitted papers to Nature, both groups reaching very similar conclusions about disease spread and the effectiveness of interventions.

An important conclusion is that the potential for containment depends largely on the reproductive number, R0, which is the mean number of secondary infections deriving from a single individual in a totally susceptible population of given structure. The nuances of this number are important. R0 must be greater than 1.0 for sustained transmission; however, it is important to recognize that in reality R0 is not only a property of the disease, but also of the social network, immune status of the population, environmental effects, and history of the pathogen. For the MIDAS models, the average R0 was about 1.4, not a high number but problematic because the transmission time is very short. The models also show that containment is sensitive to delay in implementing intervention strategies and suggest that a combination of strategies may be more effective than any one strategy.

The steering committee congratulated the research groups on their progress.

Collaborations with non-MIDAS Researchers

Irene presented a draft policy for establishing associated projects - that is, projects that are closely related to MIDAS but are supported outside of the project. The committee raised several important considerations, including ensuring high quality of projects, providing for some sort of review, and implications for computing/data resources.

The committee agreed that prospective projects should submit proposals with the MIDAS PI to the steering committee for consideration. Proposals should detail the nature of research, the nature of the collaboration, agreements for data, tool, and model sharing, and other information that may be required by NIH.

MIDAS Working Groups

Five working groups presented their plans to the steering committee (documents are available on the MIDAS Portal). These presentations were informational, but they did raise a question of how to direct and focus the groups' work to ensure that the MIDAS mission does not creep too far.

The committee recommended that the PI's be charged to meet regularly as a group to plan, coordinate, and provide direction.

The meeting adjourned at 4 pm.