See the publication suggestions also for general instructions that apply here. These are important considerations for reviewers during a grant application review based on my experience as both a writer and as a reviewer (see my CV for more details). While this by no means a guarantee of success (and conversely, some of these comments are not generally applicable), just following this as a recipe I think will put you a notch over other applications that don't.
The most important thing you can do to learn how to write grant applications is to serve on study sections and review panels. If you can't do that, then read my notes on what happens in a study section/review panel.
- Avoid using weak language like "we plan to..." or "we can...". Say "we will" whenever possible instead. Be forceful.
- Know your program officer and the goals of the program you're applying to. Look at the composition of study sections that are likely to review your application and try to determine likely reviewers. Even though you can't predict who will review the application, put yourself in their shoes as you write the application and think "if so-and-so read this application, what would their reaction be?"
- Avoid careless mistakes and proofread carefully. Typographical, spelling, grammar, and punctation errors, incorrect page numbers, poor layout and organisation will simply antagonise overworked reviewers.
- Specific aims should each consist of three sentences that explain what you're going to do, how you're going to do it, and why it is important (the latter two can be reversed but it needs to be consistent for all aims).
- Each part (what, how, and why) should be one sentence each. Deviation from this criterion is possible, but should happen only if absolutely needed to convey the appropriate information.
- The first sentence of each aim (describing what it is that will be done) should be in bold.
- Start your specific aims with action verbs, i.e., "Create...", "Develop...", etc. If you have N aims, you should have N distinct action verbs (i.e., avoid repeating the same action verb unless absolutely necessary).
Background and significance
- Compare yourself to other relevant work being performed in this area.
- Indicate what is novel with what you are doing and why you're most qualified to do it.
- End section stressing the significance of what you're doing.
- Have a LOT of preliminary results.
- The more papers you have, the better it is. Even if you have great preliminary results, people aren't going to believe you until you have published papers.
- Propose good controls for everything you do. Try hard to figure out and explain what the odds are of obtaining a particular result you expect to obtain by chance.
Research design and methods
- Depending on mechanism and budget, tailor application to be focused and narrow for research projects; expansive for program project (still should revolve around single ideas). Have a mix of mostly conservative (60-70%) and a few innovative (but not too radical) ideas, especially for the NIH.
- Have good collaborations to apply/validate/support the research findings (i.e., collaborate with experimentalists to apply computational techniques developed for validation purposes). Find a specific problem/application to work on and focus on that, instead of proposing broad solutions.
- Do not wait for the day before deadline to submit proposal. It is best if you are able to finalize the proposal ahead of deadlines and keep at least one week for paper work (like getting signatures). Do remember that people aren't always around.
- It is ideal if a senior group member (like Ram or Hong) can submit the proposal.
- Follow up with the SRA to see if you can submit an addendum that describes incremental/latest work you've done since the grant application was submitted. This shows to the reviewers that you're still serious about the project.
- If you've gotten a great score/rating, do keep in touch with the program officer and make sure everything goes smoothly.
- If you're on the edge of a payline, don't be shy about negotiating short-term intermediate funding (in NIH language, this is known as "select pay").
- Take reviewer comments seriously when you respond (even if you don't agree with them).