DataScience@NIH

Driving Discovery Through Data

0 Scooped! A very personal case for open science
/ 10.12.17

I was three years into graduate school the first time I got scooped. Or, to be more precise, the first four times I got scooped—all in the span of one month.

My project was rooted in the budding field of epitranscriptomics—the RNAophile’s take on epigenomics. I was trying to develop a high-throughput technique to detect the RNA modification pseudouridine, an isomer of the base uridine and the most ubiquitous of RNA modifications. Once an obscure modification, pseudouridine had been garnering more interest after a report suggested its presence could reprogram the genetic code. But its specific biological function was still rather mysterious.

I set out pseudouridine-hunting with a mission to map the modification throughout the transcriptome. If I could identify where pseudouridine was present—what sorts of RNA are modified and at what positions—I could better understand the biological role pseudouridine played.

I was the only person in my lab working on the project, so my advisor made sure I attended conferences and symposia where I could meet leaders in the field. Many of them told me the project was risky but worth pursuing, especially because they hadn’t heard of anyone else taking it on. I would, therefore, have the time and space to gain the biochemistry and programming expertise needed to hunt down pseudouridine.

As you can imagine, it came as a total shock when not one but four groups beat me to the punch. They had all independently developed the same technique as the one I had been quietly working on. As I was recovering from the blow and measuring my next steps, I wondered: how was it possible that had no one knew this was coming?

My third-year slump—a requisite of the graduate experience—had become a cautionary tale of what happens when science is conducted in isolation. Three years of hard work had gone down the drain at a crucial time in my development as a young researcher hoping to make my mark.

After these four methods were published, a colleague asked me to write a review of their approaches. I saw this as an opportunity to go beyond the published manuscripts and dive deep into their data and methods. All four groups had similar methods, and three of them had identified pseudouridine sites in the same budding yeast strain. I was curious: how well did their results overlap? In other words, how robust and reproducible were their methods?

I retrieved their raw sequencing files, mined their annotations, and reverse-engineered their data analysis workflows. Through this analysis, I found that of the 450 sites identified in protein-coding RNAs, only two overlapped. I offered technical and biological explanations for the variability, but the fact remained that these techniques, published in high-impact journals, would likely result in false leads and dead ends in the quest to understand the biological role of pseudouridine. As published, they were not robust enough to be replicated.

As I was analyzing these methods, I stumbled across the work of the Center for Open Science, a non-profit technology organization, in The Atlantic. I learned there was a name for what I was doing—a whole movement to build an enterprise around transparency and sharing. The Center had just published their reproducibility study, which resonated with my own efforts. Without knowing, I had walked right in to the open science movement.

The ideal of an open science enterprise is noble, justified in any number of ways as an obligation to society. But the promise of open science has a very human benefit, palpable in young researchers who’ve had the “publish or perish” mantra beaten into our brains. A culture of sharing and openness lifts many of the burdens placed on trainees to crack open the mysteries of the universe in five to seven years. It reduces the chance of ending years of research by being scooped or of publishing rushed conclusions based on a bug in our code or an artifact in our data. And it can provide trainees with the space to explore different skills, data types, and analyses with input and review from a community of collaborators.

Before graduating, I was scooped one last time. But it felt less like a crisis and more like a moment to commit myself to working toward the goal of open science. Based on my work and experiences as a graduate student, I devoted a whole chapter in my dissertation to making a case for open science.

I’m now an AAAS Science & Technology Policy Fellow in NLM’s Data Science Coordinating Unit. My job in part, is to orient myself to the great progress being made by organizations and institutions around the world to incentivize and codify a culture of sharing. As a result, I now welcome anyone and everyone looking to beat me to the chase.

---

About the Author:

Dr. Maryam Zaringhalam is currently an AAAS Science & Technology Policy Fellow at the National Library of Medicine. She received her Ph.D. in molecular biology from the Rockefeller University, where she used protozoan parasites as a model to investigate how small changes to our genetic building blocks can affect how we look and function.

Add New Comment

Posting Calendar

December 2017

Sun Mon Tue Wed Thu Fri Sat
 
 
 
 
 
1
 
2
 
3
 
4
 
5
 
6
 
7
 
8
 
9
 
10
 
11
 
12
 
13
 
14
 
15
 
16
 
17
 
18
 
19
 
20
 
21
 
22
 
23
 
24
 
25
 
26
 
27
 
28
 
29
 
30
 
31
 
 
 
 
 
 
 
Back to Top