REPOST: Structure Of An R01: Background and Significance

March 20, 2009

Reposted upon request of a dear friend of the blog.
In our first installment of a series on structuring an NIH R01 research grant application, we discussed in detail the first section: the Specific Aims. The Specific Aims page encapsulates the entire gist of the grant in one page, and if that is all a reviewer reads, they should feel all excited and jazzed about what you propose, or your grant is doomed.
In this post, we discuss the next section following the Specific Aims, the Background and Significance. The Background and Significance is designed to set the context for the proposed studies in terms of what is already known in the area of proposed inquiry, what key open question(s) are important, why they are important, and how the approach(es) of the proposed studies are highly suited to addressing the open question(s).


For a 25 page R01 application, a reasonable Background and Significance section will be about four or five pages, and will contain three or four figures. (Since the R01 is going to most likely become a 12 page application at some point in the not-too-distant future, just take the quantitative aspects of the advice given here and divide by 2.) This section will be divided into subsections, each of which should be headed with a substantively descriptive title.
Roughly the first page is used to lay out the most broadly conceived biological (or medical, in the case of non-basic-science proposals) context for the proposed studies, starting with the foundation of the entire field. So, if your application relates to some molecular/cellular hypothesis concerning the role of a particular signaling protein in growth-cone-mediated axonal guidance in the developing nervous system, you would start by laying out that the adult nervous system has a very complicated wiring structure, and in order to set that structure up, neurons in the developing nervous system need to send their axons to the right destinations. Then you would say that the growth cone is a specialized cellular structure that implements this navigational problem. And then you would talk about the basics of growth cone function broadly. So this subsection could be entitled “Axonal Guidance in the Developing Nervous System”.
It’s good to use schematics or model figures in the Background and Significance, as a picture is worth 1000 words. Also, you are permitted to use a smaller typeface in your figure legends than in the body text of the application! In this case, the first figure might show some axonal growth cones wending their way through the brain, or whatever.
Then over the next couple pages, you gradually become more specific about the current cellular and molecular understanding of how growth cones actually implement navigational decisions through the developing nervous system. Here you might have a schematic of the growth cone itself, showing the cellular and molecular components that make the growth cone work: cell adhesion and recognition molecules, cytoskeletal components, signal transduction components, etc. This gradual narrowing over these couple pages should end finally in a clear explicit statement of the unanswered question that arises out of the existing work in the literature. Like, “What is the role played by the cell adhesion molecule ZX-CAM in growth cone guidance?” And you also have to have made it very clear why this is an important question to address.
The headings of this section could include “Cellular and Molecular Basis of Growth Cone Navigation”, and the most specific final section could be “ZX-CAM Function in Growth Cone Navigation”
The next page to page-and-a-half should be devoted to providing methodological background, again illuminated with a schematic figure or two, and divided into titled subsections, one subsection for each technique. If you are using established techniques, explain them through reference to the literature, and explain how they are applicable to answering the important question you posed. If you are proposing the use of a novel technique–which, in general, is a good thing to be doing from a “sex-appeal” standpoint–then explain how the technique is envisioned to work (with a schematic) and would, in the ideal situation, be highly applicable to the important question you posed.
In either case, you refer forward to the Preliminary Studies, where you will be providing direct experimental evidence that the chosen approaches will be effective at answering your important question: “As will be described in detail in the Preliminary Studies section, we have validated the applicability of the blah technique for addressing the whatthefuckever question.”
Then, the most important subsection is the last paragraph of the Background and Significance, entitled “Summary of Background and Significance”. This subsection should be a completely self-contained summary of the entire argument of the Background and Significance, with each paragraph (roughly) encapsulated into a single sentence. So, you restate the general area of inquiry, why the specific important question to be addressed is important, what the chosen approaches are and why they are applicable.
This final Summary paragraph then ends with one or two special sentences, depending upon if you are proposing the use of a novel technique that you have developed for addressing your important question. If not, then you end with a sentence that explains why your proposed approach and model system is uniquely suited to addressing the important question, and why the results will be relevant in a broader context.
If you are proposing the use of a novel technique that you have developed, then you follow that sentence with a sentence explaining how your novel technique will also be applicable in a substantially broader context than just your particular important question.
And remember, every single word you write in this section (actually, the whole fucking grant) should be directed towards establishing that your question is important and your approaches are uniquely suited to answering it.

No Responses Yet to “REPOST: Structure Of An R01: Background and Significance”

  1. Neuro-conservative Says:

    Well said, PP, and I would again add that the subsections should not only point forward towards the preliminary studies but also back to the specific aims.
    That said, I am quite nervous about potential unintended consequences of the reduction to 12 pages:
    1) Not enough space to develop a novel argument, leading to further encouragement of safe, stale topics and approaches;
    2) Newer PIs with shorter/less-well-known published track records will be disadvantaged by needing more space in the Prelim Studies;
    3) Overburdened reviewers may not spend the extra time reading submitted appendices or cited publications;
    4) Underspecified grants will lead to a more subjective review process;
    5) Shorter applications will mean lower barriers to entry and will encourage even more applications resulting in still lower paylines — analogous to the frequently observed phenomenon of wider highways leading to greater rather than reduced traffic burdens.
    Those are just the first few pitfalls that come to mind. I hope the NIH slowly pilots this idea before full implementation.

    Like

  2. drdrA Says:

    I already submit to a federal agency that allows only 18 pages and I am squeezing every word in. Cutting down to 12 pages will be real interesting…

    Like

  3. juniorprof Says:

    Thanks for this, very useful, yet again. I am looking forward to the switch to the 12 page format. In my time in Canada I got used to the shorter format of the CIHR R01-type grant and quite liked it. Its funny, I never really imagined that that particular experience would come in handy but it obviously will. Goes to show, you never know when a variety of aspects of your training will come back to pay off.

    Like


  4. PP,
    One thing that would be useful in this series would be to touch on how one ties together the whole grant. That is, specifically tying sections in the Background, Preliminary Data, and Research Design/Methods to particular specific aims.
    On the topic of 12 pages, I share Neuro-conservative’s concerns. On the other hand, since NIH is having a hard time finding reviewers, NIH might not have any other choice.

    Like

  5. neurolover Says:

    I’d love to hear more info on NIH’s difficulty finding reviewers — as far as I can tell, the main group from which reviewers are drawn are folks who hold NIH grants. Are people difficult to find for fundamental reasons, or because it’s obviously know fun to read a gazillion grants, and admit that you just can’t fund practically any of them? That’s a systemic problem that won’t be addressed by making the grants shorter.
    I think Drugmonkey’s suggestions about putting grants in “waiting lists” would be more likely to result in more job satisfaction by the reviewers.
    But, the fundamental issue is that dividing small pots of money is just no fun at all.

    Like

  6. Beaker Says:

    A colleague once gave me the following advice about this Background and Significance: put the Significance up front. The reasoning was that the reviewer has to write something in their critique about significance, so if you write it for them and place it at the beginning, it is possible they will just parrot what you wrote, making you come out smelling like a rose.
    I have mixed feelings about this advice. I tried it, and indeed, the critiques did sound a lot like what I wrote. On the other hand, It is always better to let the reviewer think that they thought of something themselves after reading your arguments. By telling them why your research area is significant– before “showing” them with citations, logic, and rhetoric–you might be diluting your persuasiveness. Thoughts from the silverbacks?

    Like

  7. TreeFish Says:

    Though I’m not a silverback, I will quote a deceased and famous silverback, Ted Bullock. He relayed this advice for giving scientific talks, but he also abided by it for papers, and presumably grants: “Say what you’re going to say. Say it. And then say what you just said.”

    Like

  8. DrugMonkey Says:

    Neurocon: I hope the NIH slowly pilots this idea before full implementation.
    HA! fat chance. You have clearly not been paying attention to all the constant drumbeat of “wouldn’t shorter be better? wouldn’t it? how can we frame the queries to get our desired answer?”. It is going to happen and fast.
    MtM: One thing that would be useful in this series would be to touch on how one ties together the whole grant.
    I don’t know about the whole grant but one thing I’d add is that for just about every paragraph of the Background/Significance statement I think one should explicitly tie this to the research plan in some way. PP has this for the way he forwards to the Preliminary data section but everything can take this approach. Everything from the broad hypothesis opening to the discussion of specific methods can benefit from something like “..this question is to be tested in Aim I”. or “..which is why we will use these methods in Experiment 6 of Aim II”. or “these competing hypotheses will be directly tested by…”
    while you want to have the Background tell a nice story from broad interest to specific experimental approaches, do not fear to give away the ending!
    MtM: since NIH is having a hard time finding reviewers, NIH might not have any other choice.
    Neurolover: I’d love to hear more info on NIH’s difficulty finding reviewers
    I think it is an utter crock, personally. If you read what they actually say, they are talking about the “right” reviewers, the “high quality” reviewers and the like. I doubt very much they have trouble merely with warm bodies.
    My problem, as always, is that the official actions always flail around like this, create a lot heat and spin-cycle belief in the proletariat without providing a shred of backing evidence. I constantly fence with more senior (than me) scientists about this issue and they parrot the quality-of-review tropes without any specific evidence either. They clearly believe what they are saying, I just can’t get any data out of ’em or any coherent discussion at all.
    Old Scientist: “Junior reviewers suck!”
    Me: “oh and why is that? and how do you know?”
    OS: “picky….detail oriented…can’t see big picture…blah, blah”
    Me: “oh, and how much of this is simply a factor of reviewer experience, regardless of age/academic tenure? and since study sections are limited to no more than 10% asst prof (or less) and those are most likely to be ad hocs with minimal load what great breadth of experience with all kinds of reviewers do you personally have?”
    OS: “hummina, hummina, hand wave bs..bs. blah”
    it never fails.

    Like

  9. Perceval Says:

    Aww, I am so envious of your twelve pages. No, seriously – here in the UK we have six for the main proposal, two for track record, and one for justification of resources. That’s EPSRC, who cover engineering and physical sciences. Well, we do now have to do a two page impact statement as well, but nobody is sure what that is for.
    Six pages is really hard though – you are v vulnerable to “why haven’t you cited meeeee?”. Or you can get accused of not reading key conference proceedings in a proposal where nearly one of your precious six pages was devoted to references …

    Like

  10. Curt Fischer Says:

    CPP, thanks for these recent posts. They are a joy to read, and they come with so much useful information to boot! I hope posts and comment threads like these keep on coming here at DrugMonkey.

    Like

  11. cashmoney Says:

    They are a joy to read
    Reminds you that PP can write the hind legs off a donkey when s/he wants to, eh?

    Like


  12. You fuckers want me to finish the motherfucking “Structure of an R01” series, or what? I never got to “Preliminary Fucking Studies” and “Research Design and Goddamn Methods”.

    Like

  13. Beaker Says:

    CPP,
    Do a quickie on the “Preliminary Studies” if you want, but my main interest is your advice for how to write the “Research Design.” Here are two small issues and then one big one about which I’d like to hear your advice:
    Small issue #1: how important are those “letters of support” from bigwigs?
    Small issue #2: is is wise to flood the appendix or other parts of the experimental section with a detailed technical section in order to provide “reassurance” to the study section that you really are a serious scientist, who is not just making shit up?
    The Big Question: how does one best frame having a “plan” for discovery-based research? If you knew what was gonna happen, then it wouldn’t be research! Besides the no-brainers like having great controls (both positive and negative), how can a person write a 5-year plan for a whole project without writing fiction? If anybody actually does precisely what they propose for the whole grant period, then they didn’t discover anything novel, right? How much effort do you put into this sort of if/then crap:
    “If our assay does not work, then we will decide to not continue wasting our time on something that will not generate useful data.”
    “If our hypothesis is refuted, then we will decide to not continue to produce new experiments in this area, in a futile attempt to revive a dead hypothesis. Instead, we may boldly attempt something else instead.”
    “Should our first brilliant idea not work, we will attempt a slightly less brilliant (but still highly feasible, see preliminary data) alternative plan.”
    If you do offer advice, would any of it be different once the NIH goes to the 12-page application?

    Like

  14. DrugMonkey Says:

    The Big Question: how does one best frame having a “plan” for discovery-based research?
    By doing as best you can to conceal the discovery part and by inventing dry hypothesis-test-interpret structures into which to fit your ideas.
    No matter how much you (or I) might want to rail about how science is “really done”, the StockCritique of “not clearly hypothesis driven” is a reality and something you need to account for.
    Does this mean that all study sections will kill you for discovery-based research? Maybe not. But 68.675% of them will 99.5% of the time and 96.78% of them will do so at least 81.26% of the time.

    Like

  15. Beaker Says:

    By doing as best you can to conceal the discovery part and by inventing dry hypothesis-test-interpret structures into which to fit your ideas.
    If the situation is truly that dire, then I picked the wrong profession. When I say “discovery-based,” I do not mean experiments which are subject to the stock criticism of “fishing expedition.” Good discovery-based research is still hypothesis-driven, but a refutation of the hypothesis does not send the project straight down the toilet.
    One of the main purposes (and presumed strengths) of the NIH is to identify and pursue new research avenues. To best accomplish this, one asks important but experimentally tractable questions pertaining to medically relevant areas. It is assumed that some new discoveries in some of these areas may some day lead to big medical benefits, but that it is foolhardy to know ahead of time which ones will.
    I am not railing, and I don’t expect pity. I would simply point out that the current mechanisms of NIH grants and review are quite poor at optimizing the process I describe in the previous paragraph–unless people don’t actually do what they write in their grants, in which case the whole process is a farce. Indeed, the NIH system may be the worst system for allocating research money (unless you compare it to all of the others).

    Like


  16. I am working on a detailed “Structure of an R01: Research Design & Methods” that will answer all your questions, Beaker.

    Like

  17. Beaker Says:

    CPP–
    Woot! Eleventy!!!111!!!

    Like

  18. whimple Says:

    If our assay does not work, then we will decide to not continue wasting our time on something that will not generate useful data.
    This is a trap comment to make, even facetiously. If the assay is going to be important to what you are proposing, you need to have preliminary data that the assay is already working. The assay not working is not an option. Proposing to develop the assay is (in my experience) not an option. The assay WILL work, see, just like it did over here. After we get an answer from our assay then if the answer is (a) we will interpret that as (p) and then follow up with (x). Alternatively, if the answer is (b) we will interpret that as (q) and follow up with (y). You get the idea.

    Like

  19. crystaldoc Says:

    “The assay WILL work, see, just like it did over here. After we get an answer from our assay then if the answer is (a) we will interpret that as (p) and then follow up with (x). Alternatively, if the answer is (b) we will interpret that as (q) and follow up with (y). You get the idea.”
    This definitely seems like the way to go, and it is what I strive for… but I still have difficultly reconciling this approach with the reality of the scientific timeline, which in my experience, is that working out the assay takes 6 months and then following through with the actual experiment takes two weeks. And once having gone through the agony of working out the assay, you are not going to go into your office and write a grant with all the minute detail of how you’d follow up on outcomes a vs b; you are too excited to wait and are finally going to just *do* the damn experiment already and get the answer. I have yet to find a reasonable balance of enough preliminary data to absolutely prove feasibility, yet not too much to give away the results, or you’ve already proven your hypothesis at least in outline form, and the hypothesis isn’t novel enough any more.

    Like

  20. Beaker Says:

    Wimple, practically speaking, I agree with your suggestion 100%. I was overcome with facetiousness. Of course, you show your assay works already. To not do so makes one vulnerable to the “too preliminary” stock criticism. If the assay doesn’t work, the whole specific aim could be in jeopardy.
    The point I think I was trying to make is that, two years down the road, I’ll probably need to develop a new assay. I’ve developed scores of assays during my training. If I need a new one, I will make it. I see little point in obsessing about “alternative approaches” and back up plans. The track record should speak for itself, but it seldom does if you are a (so-called) “young investigator.” This contrasts with the silverbacks, who seem to be given the benefit of the doubt–even for hare-brained pie-in-the-sky shit that won’t work.
    Crystaldoc brings up a real-life situation that just adds to my cynicism. (S)he is afraid of “giving away the results.” In practice, what I see is that many successful RO1-writers simply use aim #1 to propose shit that they have already accomplished. Never seems to hurt them, so I doubt the fear of “hypothesis not novel enough any more” is something to worry about.

    Like

  21. Pinus Says:

    I know I am a little late to the part here…but I was curious how common it was to put in methodological stuff in here? my post-doc mentor never put anything like that in his, so I have not been, wondering if that is a mistake.

    Like


  22. It depends. If what you are proposing uses bog-standard methodological approaches, then there is no need. If you are proposing novel approaches, then laying out existing approaches for purposes of comparison should be done in B&S.

    Like


Leave a comment