Report of the PSI:Biology Advisory Committee

National Institute of General Medical Sciences
National Institutes of Health
December 21, 2010

"A good idea is a network." - Steven Johnson,
Where Good Ideas Come From: The Natural History of Innovation

The PSI:Biology program seeks to leverage the high-throughput technology developed during the 10 years of the Protein Structure Initiative (PSI) for the solution of important and complex biological problems through a partnership with laboratories expert in the biological systems being studied. Key to the success of these programs is networking of novel technologies to establish protein production and structure determination pipelines, networking among the high-throughput centers (HTC) to spur technology development and networking of the centers with their program partners so that the novel high-throughput technology can be innovatively applied to the biologically important and technically demanding problems that have been proposed. Lastly, but no less important, is networking of PSI:Biology as a whole with the biomedical community so that insight and technology is rapidly and broadly disseminated.

Based on the reports and discussions of the PSI:Biology kick-off pre-meeting in July 2010 and first annual meeting in December 2010, it is clear that significant progress has been made in establishing each of the above mentioned networks as well as establishing the individual partners programs. Members of the Advisory Committee who served during PSI-1 and/or -2 commented on the freshness of the meeting discussions that now include the biological partners. Each biological program and each membrane center has "hit the ground running" and is moving their work rapidly forward with their partner HTC. The PSI Knowledgebase and Materials Repository are likewise gearing up for the onslaught of information and material that will be generated by PSI:Biology. Below we briefly summarize key issues that were raised during the annual meeting and feedback the Advisory Committee offers to help insure success of the program, particularly during the critical early stages.

  • Metrics/milestones: The shift from high-throughput analysis of the structural "low hanging fruit" to difficult and complex targets defined by their biological importance mandates a change in the metrics/milestones to be used to judge the success of PSI:Biology. The issue is nicely summarized by the question "How do we measure units of biological knowledge?" posed during the annual meeting. The Advisory Committee encourages on-going discussion of this issue and endorses the plan to have each biology partner formulate a "milestones proposal" that will be the basis for further discussion among the PSI:Biology participants and ultimately implementation. Metrics/milestones being discussed include:
    • The number of structures solved that have high biological impact, such as those that resolve longstanding questions, open new areas of inquiry or force a paradigm shift;
    • The number of structures solved that were "not accessible" to PSI-1 and -2, such as multi-protein complexes, eukaryotic domains, full-length eukaryotic proteins and membrane proteins;
    • The number of clones and purified proteins provided by the HTCs to both their biology partners and the community at large;
    • The number of publications that cite the structures solved by PSI:Biology;
    • The number of novel technologies developed by the HTC and membrane pro-tein centers;
    • The ability of the HTC-biology partnerships to provide the community at large with biological insight into the problems that are being explored; and
    • Scientific use cases and user profiles should be collected from all centers and participating groups to guide the development of metrics and data standards.
  • Networking and communication: Clear and efficient communication among the constituencies of PSI:Biology is essential to the success of this initiative. The membrane centers have already self-assembled into a highly interactive group. Effective networking among the HTCs and between each HTC and its partners has also been established and is a significant contributor to the fast start of PSI:Biology. Issues that the Advisory Committee identified as needing to be addressed include:
    • A Coordinating Committee should be established to facilitate communication between the research centers, projects groups and NIGMS program officials. It is suggested that this committee be composed of representatives from the HTC, membrane centers, biology partners, PSI Knowledgebase, PSI Materials Repository and NIGMS representatives.
    • It was noted during the annual meeting that formal lines of communication among the biology partners currently do not exist. This omission is the only shortcoming to the outstanding established PSI communication network. It is anticipated this issue will be addressed by the Coordinating Committee.
  • Outreach: Outreach to the biological community has already been undertaken by the PSI Knowledgebase and individual centers and is outstanding. Keep it up!
    • New mechanisms to reach the community were discussed during the annual meeting, such as Facebook and Twitter. The Advisory Committee urges that these be explored.
    • The PSI Knowledgebase has an impressive number of features that will be useful to the broader biological community. A mass e-mailing to members of various societies announcing the knowledgebase should be considered.
    • The committee encourages the knowledgebase and individual partners to coordinate efforts to organize Web pages and database portals in such a way as to increase appeal and accessibility to a wider community, including even non-specialists and the general public.
  • Information overload and databases: There will be a greatly increased heterogeneity in the types of information acquired by PSI:Biology compared to PSI-1 and PSI-2. Furthermore, there is clear evidence that an immense amount of information will be collected. In addition to depositions to the PDB and the knowledgebase for targets, structures, homology models, publications, protocols and technologies, a large amount of raw experimental data will be collected. This information is of great value to both the PSI:Biology partners as well as the community at large. However, its storage and curation present significant challenges that must be addressed if the value of the information is to be fully realized. The Advisory Committee recommended several actions to insure that this information be optimally useful:
    • Procedures must be established to ensure the timely updating of database information and verification of its accuracy. For example, biologically important interactions/complexes of proteins/ligands should be unambiguously flagged in the PDB-format Knowledgebase files.
    • The long term provision of the PSI:Biology databases for the biology community at large is beyond the scope of this initiative. NIGMS should explore mechanisms by which the databases generated by PSI:Biology can be spun off to another organization capable of curating and managing them into the future;
    • The integration of the heterogeneous mix of structure and functional information into informative "reports" presents significant challenges. Effort within PSI:Biology should be expended towards maximizing the usefulness of the acquired information. Primary structural and functional contributions by PSI:Biology should be clearly attributed.
  • Looking forward scientifically: Molecular modeling and the generation of models has been a significant component of PSI-1 and PSI-2 and will continue with access to these models provided by the knowledgebase portal. The Advisory Committee recommends that the PSI:Biology community begin to explore the feasibility of integration of molecular modeling and molecular dynamics into the mix of studies being conducted. Dynamics is viewed as particularly important for membrane protein studies.
  • Center site visits by advisory committee: It was suggested during the Advisory Committee meeting that site visits to the PSI:Biology centers might help committee members understand and appreciate the novel capabilities of the centers.
  • Feedback loops: The Advisory Committee recommends that a process be established by which important information can be effectively shared with the PSI:Biology community in a timely manner. Various mechanisms including e-mail blasts, Web posting and tweets were discussed

From the PSI:Biology Advisory Committee