A guide to getting funded
Researchers face challenges every day. That’s what we are: problem solvers, reaching into our toolkits, exploring mechanisms, searching for the right devices. One of the toughest challenges we face in biomedical research is explaining the science behind it, making sure that the meaning is clear, and convincing others that it has the potential to improve public health. These are core components of successful grant writing – irrespective of the writer’s status, be it a graduate student or a tenured professor. For graduate students interested in pursuing predoctoral fellowships, it is (even more) important to detail the expected skills learned from conducting the proposed science.
Let’s face it: writing a grant can be a daunting task, especially for a beginner. So, here are some tips and tricks to begin writing the grant or even refining a grant for submission/revision based on my grant writing experience as a Ph.D. student.
To help explain my strategies, I apply Research Strategy excerpts from my F31 Predoctoral Fellowship and Young Scientist Research Grant from the National Institutes of Health (NIH) and QIAGEN as templates.
Grant Components
Several sections constitute an entire grant proposal, irrespective of the institute/organization for submission. With respect to the NIH structure, the more science portions of a grant include the ‘Specific Aims’ page and ‘Research Strategy’. Essentially, the ‘Specific Aims’ page is a condensed version of the ‘Research Strategy’, and therefore details in this guide geared toward the ‘Research Strategy’ also apply to the ‘Specific Aims’ page. The ‘Research Strategy’ is broken down into several sections (Fig. 1).
Figure 1. Components of the Research Strategy section
Research Strategy
Scientific Premise*
The ‘Scientific Premise’ is no more than 2–3 sentences that summarize the purpose of your research. This section should generally provide three messages: disease of interest, relevant target and proposed mechanism.
Example: We are focused on modulating the gut microbiota and their metabolites that inflict dramatic changes on host physiology and potentiate liver cancer pathogenesis. For this proposal, we posit that short-chain fatty acids and secondary bile acids are human relevant, translational therapeutic targets to reverse immunosuppression and to alleviate hepatocellular carcinoma (HCC).
Three Messages Breakdown:
- Disease of Interest: Liver Cancer (i.e., HCC)
- Relevant Target: Gut Microbiota Metabolites (i.e., short-chain fatty acids and secondary bile acids)
- Proposed Mechanism: Immunosuppression
This section would be best placed at the beginning of the ‘Research Strategy’ because it provides the grant reviewer insight into what they will be reading on the remaining pages of the proposal. It is also important to immediately front-load key terms such as ‘human relevant’, ‘clinically relevant’, ‘clinically achievable’ and ‘translational’. Overall, this section signifies a proposed solution to help resolve a public health problem. Remember, the scientific premise just touches the surface of the proposed research; so there is no need to worry about defining jargon-like terminology at this point. The next section, ‘Significance’, details deeper explanations of the background.
*Paraphrasing the scientific premise can become the project narrative of a NIH proposal.
Significance
Example:
A common thought (or rather mistake) is jam packing as much literature information into the page-limited grant as possible to sound knowledgeable about the field. The phrase “less is more” can be subtly applied when it comes to writing the background of a grant. To better prepare a succinct grant, answer the ‘W’ and ‘H’ questions: WHO, WHY, WHAT, WHICH and HOW. Start by answering the questions in a punchline format of just 2–3 sentences and then expanding with specific examples from the literature.
WHO: Information herein includes population demographics and statistics about the impact of disease on humans.
Example: HCC has emerged as a leading cause of cancer-related deaths globally and in the United States. In 2015, there were 854,000 new liver cancer cases and 810,000 deaths globally, where the highest incidences from 1990-2017 were found with male dominance, in the high-risk elderly population and a strong prevalence in developed countries. These statistics are attributed to the asymptomatic nature of HCC and subsequent limited therapeutic intervention.
WHY: Information herein can be written in two different ways. The first approach is describing the reason for conducting the research in terms of public health (see example #1). The second approach is posing your research as a question, which can help highlight the current gaps in scientific knowledge (see example #2). For both approaches, it is important to indicate the proposed target.
Example #1: The potential harmful aftereffects of current HCC therapies emphasize that new targets are warranted for patients. We propose that targeting immunosuppression mediated by gut microbiota metabolites should be a focal point in this endeavor.
Example #2: It is unclear why immune cells become negatively imbalanced to favor immunosuppression and HCC. Recent literature elucidates that gut microbiota metabolites could be a contributing mechanism.
WHAT/WHICH: Information herein includes the research targets that have yet to be investigated e.g., a specific protein, signaling pathway, cell type, bacterial species, etc.
Example: It is not known which intestinal bacteria and what gut metabolites could promote tumor progression. Based on the literature and our preliminary data, we propose that microbial-derived short-chain fatty acids and secondary bile acids could be promoting HCC via immunosuppression.
HOW: Information herein can be written in two different ways. The first approach is describing the technical procedures for conducting the research. However, the technical procedures are better saved for the ‘Technical Innovation’ and ‘Experimental Outline’ sections of the grant rather than in the background (unless the proposal is utilizing brand-new technology). The second approach is posing your research to answer the unknown molecular mechanisms for how the biological system becomes dysregulated (see example #1) and/or how the biological system can be repaired (see example #2).
Example #1: Based on the literature and our preliminary data, we hypothesize that short-chain fatty acids and secondary bile acids promote immunosuppression by diminishing invariant natural killer T cells (iNKT) and expanding regulatory T cells (Tregs).
Example #2: Restoration of iNKT cells and suppression of Tregs are proposed mechanisms in the literature to reboot tumor immunosurveillance and abate HCC. Our preliminary data suggest that targeting gut microbiota metabolites could be an avenue to reverse this immunosuppressive phenotype and alleviate liver cancer.
Answering these questions helps to build your central hypothesis (see example #1) and summarize the goals of the proposal (see example #2):
Example #1: Targeting gut microbiota-mediated immunosuppression would halt HCC pathogenesis.
Example #2: For this proposal, we want to address which intestinal bacteria are pathogenic in HCC, what gut metabolites could aggravate liver cancer and how those microbial products potentiate hepatocarcinogenesis.
You may have noticed a pattern. My examples have many overlapping points despite answering different questions. This is normal. In fact, repetition (within reason) of certain facts and hypotheses is helpful throughout the entire grant. A reviewer may have limited time and only skim through the proposal. Thus repeating the essential information ensures that the reviewer will read the take-home messages. Apply the information priority rule: It is more effective to repeat important, relevant information than to put in excessive, non-relevant information.
Following the ‘Background’, which is literature-based, next is to highlight the ‘Rigor of Prior Research’ and ‘Preliminary Data’. This can be categorized into related prior publications from your lab and/or collaborators’ labs and unpublished data. For established investigators, more weight is given to unpublished data and should be the dominant contributor for this section. For graduate students, it is not a disadvantage to have little to no self-generated preliminary data and predominately highlight published lab material. Incorporating data that you generated is a bonus, though, because it highlights motivation, which will impress the reviewer. Remember, training grants are focused on the feasibility and validity of your experimental outline and your knowledge to troubleshoot (discussed later). As such, make sure that your preliminary data (irrespective of published or unpublished) supports the platform of your ‘Aims’. Apply the information priority rule, having unnecessary data will distract the reviewer. The preliminary data should help establish your model of choice (good for rigor) and visually display the rationale behind the target and/or mechanism of interest. This section is written in a style similar to the ‘Results’ section of a scientific manuscript.
Innovation
This section lists the novelties of the proposed research, which can be categorized into ‘conceptual’ and ‘technical’ innovation.
- Conceptual: Ideas include bridging two fields together, investigating a protein or signaling mechanism not yet studied in a field, etc. The phrase ‘… is an underexplored area of research” should be incorporated.
- Technical: Include new in vitro or in vivo models, drug candidate(s), computer program/software, advanced technology being designed or used (e.g., single-cell RNA sequencing), etc.
Approach
Rigor, Reproducibility and Statistical Analysis: Before writing the ‘Aims’ and their experimental procedures, introduce the sub-section ‘Rigor, Reproducibility and Statistical Analysis’. This is A MUST, especially for NIH grants. This is where the sample size and description of the sample size calculations for animal studies are mentioned. The specific statistical analysis approaches should also be listed.
Example: To calculate sample size for each experiment, we use a power analysis and appropriate statistical test in G*Power 3.1 software. We set experiment studies to have a power value of 0.8 and a significance of 0.05, which typically results in an experimental mouse group size to be n=5–7. Statistical significance between two groups will be analyzed using unpaired, two-tailed t-test. Data from more than two groups will be compared using one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison tests.
Student Training (Optional): For graduate students applying for a predoctoral fellowship, this is not required but I would strongly recommend including another sub-section called ‘Student Training’ before describing the ‘Aims’. As predoctoral fellowships are training grants, the reviewer strongly cares about how and what the student will learn if given the award. Hence, list the specific technical skill or skills (e.g., flow cytometry) and indicate the individual who will provide the training (ideally someone who provided a letter of support).
Aims: We have now reached the allegedly most intense (and ‘scariest’) portion of the ‘Research Strategy’: The ‘Aims’. For a NIH predoctoral fellowship, the rule is to have two related but independent aims, and this format is generally applicable to other organizations as well. Many graduate students (and even established investigators) get tripped on making the aims not dependent on each other. Meaning, aim 2 must still be a viable project even if aim 1 were to fail. As such, one method to help reach aim independence is to have an ‘umbrella’ target that can be categorized into ‘candidates’. These candidates can then be studied in your disease of interest and either have the same or different mechanism to how they contribute to the pathology. Following my example, the ‘umbrella’ is gut microbiota and the ‘candidates’ are the microbial metabolites short-chain fatty acids (Aim 1) and secondary bile acids (Aim 2).
After solidifying independent aims, the next step is to choose the appropriate experimental approaches that are specific to each aim. Each experiment is its own sub-aim and (for a predoctoral fellowship) there are two experiments per aim.
Bad Experimental Procedure: To target short-chain fatty acids, broad-spectrum antibiotics will be administered to deplete all gut microbiota species.
This is considered a ‘bad’ experimental procedure because the method completely removes the ‘umbrella’ and therefore, the results from those experiments would inform about the role of gut microbiota only in the disease context (which would be better suited as preliminary data). It won’t provide information about the specific role of short-chain fatty acids.
Good Experimental Procedure: Beta-acids will be administered to specifically deplete short-chain fatty acid-synthesizing gut bacteria members.
When next writing the ‘Expected Outcomes’ sub-section, it is important not only to write the obvious (and be specific about it) but other factors that could cause unexpected observations. This leads you to explaining the potential pitfalls and the contingency plans. These last two sections hold weight for predoctoral fellowships because the reviewer wants to observe that the applicant can readily troubleshoot. This could include either conceptual or technical pitfalls.
Example: While we do not expect any technical pitfalls with Beta-acids administration, we may find a conceptual pitfall to determine whether acetate or butyrate is the dominant short-chain fatty acid in mediating immunosuppression.
Therefore, a contingency plan would need to demonstrate how to overcome this conceptual pitfall.
Example: As an additional experiment, we would administer metronidazole to eliminate only butyrate-producing gut bacteria.
Leave some space at the end of the ‘Research Strategy’ to add in one more section called ‘Overall Goal and Training Benefits’. Reiterating to graduate students about applying for a training grant, continually emphasize how being awarded the fellowship will help you grow as a scientist, both intellectually and practically, through the expansion of your technical skillset.
Example: The proposal fits within the NRSA F31 objective to provide predoctoral individuals with supervised research training in specified health and health-related areas. Successful completion of this project will advance my scientific knowledge and technical skills, which will be pertinent for my PhD thesis and future academic career.
Additional Considerations
What you have read is the structural formality of writing a grant. Even when a template is provided, writing the dense text can still be overwhelming. To help lessen the headache, take the time to read and become familiar with the literature. Knowing what has been done can help
1) prioritize what to explain in the background,
2) identify relevant techniques to input for the methodology, and
3) recognize troubleshooting ideas.
Besides reading, one of the best ways to improve grant writing is to have others, especially those who are not experts in your field, read through a draft of your grant and take in their constructive feedback. Overall, expect many rounds of re-writing because it is impossible to write a perfect grant on the first attempt. While writing a grant takes a tremendous amount of effort, receiving a fundable score is euphoric. I hope this blog is useful to those who are pursuing grant writing. Good luck!
