10 (well, 9) ways to ensure your next grant application fails

Got this in an email from the Bioinformatics Core at the Genome Center. I saw a lot of these mistakes in an NSF Biodiversity panel that I served on earlier this year. One thing I disagree with is the notion that you cannot do bioinformatics on a personal computer. I do this all the time. Nevertheless, this is a great resource for the UC Davis crowd.

10 ways to ensure your next grant application fails:

1. Budget for data creation (e.g. sequencing) but not bioinformatics (analysis, storage, backup, dissemination)
2. Propose bioinformatics experiments that cannot answer your question
3. Design experiments with insufficient statistical power
4. Use modern omics and bioinformatics jargon incorrectly
5. Assign critical analysis procedures to untrained personnel
6. Propose to employ out-dated techniques
7. Explain your methods vaguely without attention to detail
8. Trivialize your study by not linking it to other information
9. Attempt to perform bioinformatics analyses on personal computers
10. Propose writing novel software without employing a professional programmer


The NIH RO1 deadline (among others) is fast approaching and in the month of September, grant planning in collaboration with the UCD Bioinformatics Core is 100% free*. Our typical grant preparation packet includes:

* A face-to-face meeting with the Ph.D. scientists of the Bioinformatics Core
* A letter of support from Dr. Ian Korf, Associate Director for Bioinformatics at the Genome Center, expressing our delight and enthusiasm for your awesome project
* A bioinformatics methods section written by experienced Bioinformatics Core analysts
* Figures to help your reviewers understand the importance of your study
* Cost estimates for anything from pilot studies to long-term projects
* Thorough inspection of experimental design by a professional statistician

Granting agencies are increasingly looking for proof of bioinformatics competency in awarding funds. Applications that don’t demonstrate bioinformatics capabilities within the labs of collaborators, or through other research facilities, will not score well. This includes the running of standard and innovative bioinformatics analysis pipelines as well as design, purchase, and maintenance of the requisite high performance computing hardware. The Bioinformatics Core at the UC Davis Genome Center has 10 years of experience analyzing large biological datasets, from microarrays through Illumina and PacBio sequences, in close coordination with the Genome Center’s other experimental core facilities. We have 6 years of experience teaching bioinformatics analysis to grad students, postdocs, and PIs. And we have 10 years of experience designing and running high performance computing resources, including ROCKS clusters and high-memory servers.

Contact us at bioinformatics.core@ucdavis.edu, or 530-752-2698, and for more information, visit our website at: http://bioinformatics.ucdavis.edu/

* Actually, it’s always free to talk with us at the beginning of a project. We’re here to help!

One thought on “10 (well, 9) ways to ensure your next grant application fails

  1. While I also have done bioinformatics on a desktop computer or laptop (normally >8GB RAM), I would NOT recommend leaning heavily on this in your grant application. Normally personal computers are OK for data parsing (converting/filtering files), low memory tasks (e.g. closed-reference OTU picking), or data visualization (R code); during the course of most funded projects you will need to utilize a high-memory server or run jobs in parallel on a server. In a grant, I would say that it is fine to mention lab/office computing facilities will be utilized for small tasks, but proposals need to have a significant provision for bioinformatics resources (either cloud-based or in collaboration with a core facility or high-performance computing group). Otherwise the assumption is that you won’t be doing cutting edge science.

    Like Jenna, I have also seen MANY of these mistakes in grant proposals I’ve reviewed for various NSF panels during the past two years – particularly #6 (outdated technology – because 454 is now considered completely obsolete!).

Leave a Reply

Jenna Lang

Jenna Lang is a postdoctorate associate in Jonathan Eisen's lab and studies the seagrass microbiome.