As I mentioned last month, electronic submission of NIH grant proposals has already started for certain programs. The first electronic SBIR/STTR applications were due several months ago. AREA proposals (a.k.a. R15s), which support research at smaller and less research-intensive institutions, made the switch in late February. "Small grants" (R03s) and "Exploratory/Developmental" grants (R23s, etc.) will switch over 1 June. For R01s, the first electronic-submission deadline will be June 2007, and NRSAs will follow in August. For the complete schedule, consult NIH's complete (and updated) transition plan .
Once a program makes the transition to electronic submission, paper proposals will no longer be accepted. Converting to the new process is not rocket science, but it's not trivial either. It requires some effort from you and your institution, so you need to be prepared.
If the process seems more complex than it ought to be, it's because NIH isn't merely switching to electronic application forms; they have a legal mandate to switch to a government-wide standardized electronic-submission procedure. Specifically, NIH's electronic-submission program is a collaboration between eRACommons (NIH's electronic research administration organization) and Grants.gov, which is managed by the Department of Health and Human Services. Both organizations--NIH and Grants.gov--have standards, and electronic applicants must meet both sets of standards.
The process of submitting electronically is iterative and takes time, so get started early. First, both you and your institution's "Authorized Organizational Representative/Signing Officer" (AOR/SO) must register with eRACommons and Grants.gov. NIH has done an excellent job documenting the registration process, with detailed instructions  and a training demo .
For your AOR/SO, registration might be more complicated. Before they can register with Grants.gov, they first have to register with Central Contractor Registration  (CCR) and obtain a Data Universal Number System  (DUNS) number.
Once you and your AOR/SO are registered, you prepare an application using the new SF424 (R&R) form in place of the old PHS 398 (the application guide  is downloadable of course), and your AOR/SO submits the proposal one piece at a time. Errors will be generated during the submission process (these are technical errors generated by Grants.gov), and here's a point of near-universal criticism of the process: It is not possible to display and print a comprehensive list of errors; the errors are listed one at a time.
Once the errors have been fixed and the application has been submitted in (technically) error-free form, it's NIH's turn. NIH probably will find a problem or two, so they won't validate the application immediately. Instead, they will return it with comments. Once NIH finds the application acceptable, your AOR/SO must sign it and submit it once more, and the application is validated. The submission process is complete, and the review process starts.
NIH says that you need to begin the process well in advance, at least a month before the deadline or before you intend to submit your proposal. Each application/feedback iteration can take up to 2 days. You must register 2 weeks before the deadline, or NIH won't consider it a "good faith effort" and you may be penalized for any technical delays. Technical errors should be fixed before the application deadline (8 p.m.-- not midnight--on the application due date), but the validation process can continue to 1 week past the deadline.
As in any project of such complexity, a lot can go wrong. To address the inevitable problems, NIH has put together an extensive FAQ . Here is a list of some of the problems applicants have faced so far, drawn from personal interviews and NIH's own documentation.
The oldest advice for anything computer-related is still the best: Save your work often. Walter F. Horton, associate director of the School of Biomedical Sciences at Kent State University and director of graduate studies at Northeastern Ohio Universities College of Medicine, lost a finished application when the software failed. He had to reassemble the application "from bits and pieces. I learned my lesson," he says. From then on, he "saved the electronic version at multiple steps along the way."
The system is not very Macintosh friendly. As Science reported last week , the contractor that developed the system, PureEdge Soutions, focused on the Windows platform and assumed that people using Macintosh computers would cope via Windows emulators. That approach didn't work very well, which set the schedule back by several months. NIH is working on this, but for now it's best to stick to Windows.
All submitted documents must be in pdf format, something the directions don't make clear, according to Horton. The pdf files you submit must have all security features turned off and should not include live links.
In the "Credential" field (under "Senior/Key Personnel Profiles"), it is tempting to record, well, your credentials. But if you do that, your application will be rejected, for this is where your eRA Commons ID goes. This information is required, although the form doesn't say so. Doug Wilkerson, associate vice president for research at the Medical College of Ohio, made this mistake the first time through.
Horton also mentions that the process used to submit your research plan, in which one section is submitted at a time, gives applicants less control over the total page count. "It does not necessarily use the page spacing efficiently," says Horton, "and also adds in some extra pages which then get counted against you in your total pages." Horton adds that this is something NIH is likely to take into account in determining whether you meet the page-length restrictions.
In Box 5 of the SF424 (R&R) application, you are asked to enter your institution's DUNS number, but that information might not be on file with eRACommons and Grants.gov. If the system doesn't recognize the number you enter, you'll get an error, even if the number you enter is valid. If you get this error, check with your AOR/SO to make sure a valid DUNS number is on file with both the eRACommons and Grants.gov.
In case other issues arise--and they will--Grants.gov and NIH's eRACommons have each set up their own tech-support desks. Each focuses on a different part of the process; Grants.gov answers technical questions, whereas the eRA Commons help desk focuses on validation issues. Information on contacting tech support, and when to call whom, is available online .
Reportedly, the eRA Commons help desk was hard to reach as submission deadlines approached even for the low-volume deadlines that have passed already. No matter how much NIH prepares, the situation will not be better as the first R01 deadline approaches and the system is required to deal with thousands, instead of hundreds, of confused applicants. "God only knows what will happen when a LARGE-volume deadline (e.g., the R01 deadline next February) hits Grants.Gov and NIHeRACommons," says Wilkerson. So start early. Wilkerson reports that his institution's first attempt, an AREA grant submitted for the 25 February deadline, went smoothly, with only minor hitches. He offers two pieces of advice: "1) Follow instructions carefully and 2) START EARLY!"
Pamela Plotkin submitted several AREA grants in time for the 25 February deadline on behalf of East Tennessee State University, where she is assistant vice president for research and director of sponsored programs. "It was not smooth," she says, "but we learned a lot in the process and are going to be better prepared next time around. Hopefully the NIH will be as well."
The GrantDoctor is here to help! Send questions to firstname.lastname@example.org . Please put 'GDR' in the subject header of your E-mail.
Comments, suggestions? Please send your feedback to our editor .