Writing a research proposal requires that you identify an interesting topic, spend lots of time learning about it, and then design some clever experiments to advance the field. It also requires that you articulate your ideas so any reader is convinced of your expertise, your creativity, and the significance of your findings, should you have the opportunity to carry out the experiments you’ve proposed. To begin you must identify your research question. This may be the hardest part and the most fun.
Consider the following when research possible topics:
- Your interest in the topic.
- The availability of good background information.
- Your likelihood of successfully advancing current understanding.
- The possibility of advancing foundational technologies or finding practical applications.
- The likelihood that your proposal could be carried out in a reasonable amount of time and with non-infinite resources.
Once you have a topic, consider these general tips for idea development (early and late stage):
- Read a recent review paper. This approach is one of the quickest ways to develop sufficient expertise as well as an eye for existing research gaps.
- Talk to faculty who are experts in the subject area. This is more fun and interactive than reading a review paper for honing in on research gaps! But don't expect them to do all the work for you. The conversation will be more effective if you do some research in advance.
- In later stages, faculty may also be a resource for appropriate methods to use: How do people usually measure X? Is there an animal model for Y?
- If a recently published finding inspires you, read the discussion section carefully. What do the authors suggest with respect to future work?
- Give yourself enough time (and sleep) to let your creativity emerge!
A few 20.109-specific ground rules for choosing a topic:
- You should not propose any research question that has been the subject of your UROP or research experience outside of 20.109. This proposal must be original.
- You should keep in mind that this proposal will be presented to the class, so try to limit your scope to an idea that can be convincingly presented in a 12 minute oral presentation.
You will complete this assignment in partners.
As you prepare your assignment be sure to review the resources provided on the Communication tab. In addition, see this provided example for additional guidance on layout and organization. Lastly, review the Research proposal presentation Evaluation rubric (linked here)!
Please submit your completed Research proposal slides 1 hr prior to your scheduled laboratory session time to Stellar, with filename TeamColor_LabSection_RP.doc (for example, Rainbow_TR_RP.doc).
Formatting and length guidelines
Prepare a 12 minute PowerPoint talk that describes the research question you have identified, how you propose to study the question, and what you hope to learn. A general outline for your research proposal presentation is:
- a brief project overview (scientific and social context)
- sufficient background information for everyone to understand your proposal
- a statement of the research problem and goals (specific research aims)
- project details and methods
- predicted outcomes if everything goes according to plan, and if nothing does
- special resources to complete the work (integrated when relevant)
- societal impact
You can format your slides as you see fit (keep in mind the 'best practices' tips and feedback); however, you must include slide numbers on the bottom of every slide. This is helpful in referencing specific slides during the question / answer portion of your presentation.
Your research proposal presentation will be graded by Dr. Noreen Lyell, Dr. Leslie McClain, and Dr. Becky Meyer.
Q: How '109-like' does it need to be?
A: Your proposal does not need to directly involve research pertaining to DNA damage, synthetic biology or small molecules! However, it does need to include a combination of engineering and biology, using at least some techniques / approaches that you have learned in class. For example, determining the role of a specific protein in apoptosis is not a good project (too much biology) -- but engineering a protein sensor to quantify the activity of the protein under different conditions and using that data to predict apoptosis would be an acceptable project.
In general, if you have question about whether your idea satisfies this criteria, ask an Instructor!
Q: What is an appropriate scope?
A: A typical proposal should span a few years, but some stages should be described in more detail than others.
You may not have a good feel for what a three-year project should encompass. It's not important here that you be precise. Rather, you should make sure your project wouldn't span only a few months… or over a decade. If you're still not sure, ask us directly.
We would like you to demonstrate both some breath and some depth in this proposal. For example, if you propose molecular and/or in vitro work, you would want to be able to say a few words about in vivo experiments, beyond the words “in vivo” – such as being aware of a useful model system. As for depth, you should define a few experiments in detail. What controls would you use? What potential outcomes do you expect, and how would you deal with the different possibilities? If your sample looks like your negative control, how would you modify your approach? If the sample looks pretty good but not great, how might you tweak your design? If you can't get good validation data at all, is there an alternative method to try, perhaps one that is more time consuming but more sensitive? If your synthesis fails, do you have a back-up plan? Keep in mind the outcomes matrices we made for experiments like transformation after ligation.
Q: What does novel mean?
A: We don't expect you to be Angela Belcher. But we do expect you to go beyond a direct analogue of a single existing publication.
Like a few other things in life: we know it when we see it. The level of novelty we expect in 20.109 is difficult to define precisely; it may be easiest to do so by example. You might find a useful method or interesting device that was recently published, and apply it to a totally different context, which would likely require some novel engineering design choices. You might combine improvements/approaches from multiple papers in a single area, but in novel ways that require you to make unique choices with respect to experimental set-up and validation ("add molecule X from paper 1 and molecule Y from paper 2 at the same time" is not okay, but "co-deliver drug fusion from pH-sensitive nanoparticle" might be). You might investigate a mechanism of why something worked the way it did in existing research. Or you might genetically engineer a novel biological device.
Briefly, you should be able to define a specific research gap, and a plan to address it that includes some novel methodological and/or synthetic elements.
Q: How soon should we settle on an idea?
A: Depends... on your interests, strengths, and work habits.
There is a trade-off here that you can keep in mind. You might spend weeks coming up with a great, truly novel idea, and then find yourself with too little time to prepare your slides and polish your delivery or perhaps even to define your methods in sufficient detail. On the other hand, you don't want to wed yourself to an early idea that turns out to be flawed or boring -- to you! It is normal to change your idea one or more times. If you settle on an idea very late, you'll have a challenge ahead of you. However, experience shows that it is better to present a project that you are passionate and knowledgable about than one that you have come to resent.
Q: What is "overview" versus "problem and goals"?
A: The overview is a brief, holistic summary of the project, while the problems and goals define specific stages of the research.
As the rubric (PDF download) states, one way to approach the overview is to describe both social and scientific context. Working with a previous year's example, the social context was the need for improved treatment for diseases of the cornea due to a shortage of donor tissue, while the scientific context was overcoming limitations of corneal culture -- such as apoptosis -- that lead to poor wound healing.
The problems and goals should be more specific, but still somewhat high-level. Taking Module 1 as an example, aims would be something like "Prepare truncated GFP constructs," "Validate flow cytometry assay in a cell line," and "Test recombination frequency under various conditions," NOT "Digest backbone and insert," "Ligate construct," "Transfect into cells," etc. A typical project will have three or four high-level aims.
Q: How specific should we be about "resources"?
A: Read the rubric :) We are not looking for an itemized spreadsheet or a dedicated slide. Do specify unique elements where you can.
Q: Can our proposal be purely computational?
A: This is a hard question -- and one that often falls under the 'Is this 109-like?' category. In general, you should not propose a purely computation project -- this isn't 20.320. However, here are a couple examples that do satisfy the 20.109 requirement:
- The Bidkhori et al model used in Mod2 F13 did not account for receptor trafficking. Perhaps you want to create a more realistic model of NSCLC signaling pathways. What experiments would need to be conducted (and how would they be controlled and quantified in the correct way) in order to parametrize the receptor trafficking? Is it necessary to account for other receptors? How would you do that experimentally?
- The Cancer Cell Line Encyclopedia contains an enormous amount of publicly available data about cell proliferation. What if you are interested in cell migration? Are there modeling techniques that could be applied to that data set along with experimental techniques that you could perform to harness that data and learn how cell migration is controlled?
In short, you must have some experimental component to your research proposal. You may use experiments to inform a model or a model to inform experiments -- but the experiments must be a large focus.