Conclusion of the ERCC Projects

07 Apr 2019 » James Diao » Boston, MA


After wrapping up two of my undergraduate research projects (on an analysis pipeline and data resource), I’ve been thinking about how much went into them: how much work I’ve put in and how much I’ve learned. Working in Gerstein Lab has taught me a ton about bioinformatics research, but also about the process, pace, and zeitgeist of academia.

Timeline of my two projects in Gerstein Lab

Gerstein Lab Timeline

Acronyms: ISEV, International Society on Extracellular Vesicles (conference); NESS, New England Science Symposium; YURA/YURC: Yale Undergraduate Research Association/Conference; HMS M1, Harvard Medical School, year 1

It took around 1 year to get results worth talking about. The next 1.5 years involved a lot of validation, fine-tuning, and updating (the exRNA Atlas resource and pipeline requirements both grew rapidly). In that time, we also put together abstracts and presentations to share our work with the exRNA community and get their feedback. Only then were we able to write and submit a manuscript and begin the lengthy peer review process.


In undergrad, I split my time between two labs: clinical genetics with Zak Lab and bioinformatics with Gerstein Lab. Although I spent at least as much time on and learned just as much from the former (if not more), I’m going to focus on the latter for now.

Aside from some not-very-productive wet lab work in high school, Gerstein Lab was my first serious exposure to research. I worked there for almost all of undergrad–starting freshman summer, all the way up to my senior thesis in Statistics & Data Science. It’s been interesting to reflect on how much has happened throughout the course of this project. Just looking back on our 2017 and 2018 abstracts in ISEV is enough to show how far the research has come.

Joining the Lab

I first heard about the Gerstein Lab through Jason Liu, a Yale student whom I met at a summer camp before high school. He reached out when I was admitted and talked about some of the things he was involved in. When he showed me some of his bioinformatics work on his laptop, I was so impressed (honestly, a large part from how cool the command-line looked and how fast he could tab-complete through directories). When it came time to decide on a summer internship, research in Gerstein Lab was the only job I considered. Sometimes I think about how my life would have been different if I had never gone to that math camp and met Jason. Would I have ever found my way to bioinformatics, or chosen to major in Statistics & Data Science? Would I have spent that time on bench work or epidemiology instead? Who knows! Either way, it’s interesting to think about how my life direction may have pivoted off of this one connection.

The long and rocky road to publishing

When the first papers published using our tools started to emerge, my mentor told me that the lab was going to publish a manuscript on the pipeline, and that I could contribute to its writing as a co-author. At that point, I had already set my heart on exploring more clinical informatics research (enter Zak Lab), but it was wonderfully exciting news, and I eagerly kept tabs on the progress of the manuscript and contributed to updates and edits as needed.

Over the next few years, various issues came up. Rob (the original first-author) left the lab to join a biotech startup. His new job kept him busy and the momentum to publish stalled. On our end, other ERCC projects (such as the exRNA Atlas) began to mature and take priority. It wasn’t until late 2017 (now 2.5 years into the project) that my mentor wrote up the exceRpt paper, and that our collaborators at Baylor College of Medicine began rallying co-authors for the exRNA Atlas paper.

But even with the completed manuscripts, the publication date was restricted by another factor. The ERCC intended to be submit its major papers as a “publication package” to the same family of journals. The full package of 18 papers, when finally published, looked like this:

ERCC Publication Package

Coordinating preparation for so many papers by hundreds of co-authors across the multi-institutional consortium proved to be challenging. Peer review (and the additional work requested by reviewers) added even more time. By the time the papers came out, it had been almost exactly 4 years since I started everything.

Obligatory complaining about Reviewer 3

It’s no secret that peer reviewers often disagree on their evaluations. There are countless memes of this on grad school memes with relatable themes. Here are two of my favorites:

Reviewer 3 Comic Venn Diagram Reviewer 3 Comic Monster

Exhibit A: Reviewer responses to our resubmission of the exRNA Atlas paper

Reviewer 3 Comments

*Followed by a page and a half of criticisms

Concluding thoughts

I’ve since moved on from molecular bioinformatics work, but this project was an invaluable first step into computational research and data analysis. Besides technical skills, it taught me the importance of the non-scientific part of science: all of the writing, presenting, defending, and editing that goes into convincing others that your work is good.

Although this path to publication was rocky and long (and I had definitely given up on it a couple times), I’m grateful that everything ended up working out. I’m proud of the work I put in, and I think it’s a good thing for science and reproducibility that the details of these resources are peer-reviewed and widely distributed.

<< Back to Archives