Videos Can Help Ensure Reproducibility of Scientific Results

View PDF

Biological experiments must be performed correctly. A textual method description doesn’t always capture the myriad techniques involved in even a fairly simple study. So the inability of researchers to reproduce published results is becoming a problem. And it is being addressed by video methods that show exactly how scientific results are achieved, helping future researchers learn new techniques and replicate scientific results.

The ability to reproduce and confirm a study’s results and conclusions is a foundation of scientific research. And the pharmaceutical industry relies heavily on research to identify early drug projects. Drug manufacturers spend millions of dollars conducting human trials based on data published in peer-reviewed scientific journals. But that money could be wasted because of an inability to reproduce experimental results.

In a December 2011 story by Gautam Naik (“Scientists’ Elusive Goal: Reproducing Study Results”), The Wall Street Journal highlighted a case in which a research team published a study describing how cancer tumors were destroyed by targeting the STK33 protein. Amgen Inc. then attempted to replicate that experiment in hopes of creating a drug to target the protein. After six months, Amgen’s research team was unable to replicate the study. Glenn Begley (Amgen’s vice president of research) told Naik, “More often than not, we are unable to reproduce findings’ published by researchers in journals.”

Every bench scientist is familiar with the difficulty of reproducing complex biological experiments based on text descriptions in traditional scientific journals. The text format requires authors to explain numerous small details of an experiment in a typically brief methods description. Often it is difficult to explain the many seemingly insignificant details of an experiment that must be duplicated to reproduce its results. Even if they could, experimenters would probably not bother to document many nuances and tricks that are personal habits or established practices at their particular laboratory.

Even for experienced researchers, such details can mean the difference between success and failure. The odds can be against someone who is less experienced. This both hinders research and casts doubt on some published studies. The difficulty in communicating methods can also waste time and money. Scientists can spend months trying to reproduce important findings. A portion of most biomedical research budgets (e.g., the US$31 billion annual budget of the National Institutes of Health) inevitably subsidizes the time-consuming process of learning and relearning research techniques rather than exploring new breakthroughs.

Emergence of the Video Journal

Moshe Pritsker, cofounder of the Journal of Visualized Experiments (JoVE), got first-hand experience with these

phenomena as a PhD molecular biology student working on stem cells and bioinformatics at Princeton. Once he was asked to recreate a method of culturing embryonic stem cells that had been reported by researchers in Edinburgh, Scotland. He tried following the steps in the published article but could not get the experiment to work. So his laboratory flew him across the Atlantic to spend two weeks with the group who had invented the method so they could demonstrate it to him personally. The first time he watched them, he understood what he was doing wrong, and he was able to replicate the experiment from that point on. And he recognized that the trip would not have been necessary if he had been able to watch a video of the procedure in the first place.

Pritsker came up with the idea of creating a journal to provide step-by-step video demonstrations of experimental techniques and procedures while he was at Princeton. After postdoctoral training at Harvard Medical School, Pritsker joined two friends (Nikita Bernstein and Klaus Korak) to found JoVE at the end of 2006. It presents detailed descriptions of advanced research methods in videos that are filmed by an international network of trained camera operators. The results are called video-articles or video-protocols. They allow a scientific audience to visualize experimental techniques and procedures so that small details are instantly understandable. Such visualization improves transparency and reproducibility of biological experiments and reduces the time and expense required to learn new techniques.

JoVE’s focus is more on method applicability and clarity than novelty. Its ultimate goal is to build an all-encompassing online library of video-articles and protocols for all common but complex experimental techniques in biological and medical research. They have proven to be an effective method of describing experimental techniques in biological and clinical research and pharmaceutical processing. Users can discuss and comment on articles on a video’s Web page. Numerous comments from biomedical scientists and students are providing strong evidence of the success of this innovative approach.

Videos of research techniques can help communicate the kind of details that are rarely included in methods papers but can be critical to reproducing experimental results. Video methods clearly communicate the nuances that help ensure reproducibility as well as success for inexperienced researchers or those who are switching fields.

About the Author

Author Details
Jerry Fireman is president of Structured Information, 27 Pine Grove Street, Needham Heights, MA 02494; 1-781-728-5068; jerry_fireman@strucinfo.com. Cheryl Scott is senior technical editor of BioProcess International.

Leave a Reply