blog

Recently, Cesar Berrios hosted a very successful tweetchat about diversity among researchers. You can read the summary of that discussion here. Tomorrow we’re hosting another #F1000Talks tweetchat. This time, I will be running the @F1000Research Twitter account to talk about reproducibility in research.

Reproducibility is an ongoing concern in the life sciences. Experimental results are published that later turn out to be impossible to reproduce by other groups, and it’s difficult to distinguish reproducible from irreproducible work. If we want to put more faith in work that has been independently reproduced by others, we need to change the current reward system for research, but how?

 

We have three guests lined up to discuss these issues:

Ivan Oransky (@ivanoransky) has a background in medicine and is VP and Global Editorial Director of MedPage Today as well as Adjunct Faculty at NYU’s Science, Health and Environmental Reporting Program. He also runs two very popular blogs that address issues in science reporting and science publishing: Embargo Watch and Retraction Watch. The latter blog discusses issues surrounding retraction of scientific papers.

Elizabeth Iorns (@elizabethiorns) is a breast cancer scientist and co-founder and CEO of Science Exchange, an online marketplace for science experiments. With Science Exchange, Elizabeth is also part of the Reproducibility Initiative, a collaborative effort to identify and reward high quality reproducible research.

Christin Chong (@christinchong) is a postdoctoral researcher in neuroscience, and co-founder of MicroPub, a platform for helping scientists replicate published data and incentivising reproducible science.

If you want to see what they have to say, ask them some questions yourself, or otherwise join the discussion on reproducibility in research, please follow the @F1000Research on Twitter and keep an eye on the #F1000Talks hashtag. The tweetchat starts tomorrow, April 3, at 1PM EST.

 

For some background reading about reproducibility in research, I recommend the following:

Trouble at the lab. A much-discussed feature about the lack of reproducibilty in scientific research, published in the Economist last October.

Repeatability of published microarray gene expression analyses. A 2009 Nature Genetics paper by John Ioannidis (behind a paywall, sorry!) that shows that most published microarray experiments in Nature Genetics were not reproducible, and that the main culprit was lack of available data.

Why most published research findings are false. Another well-known and well-cited John Ioannidis paper about reproducibility, published in PLOS Medicine in 2005.

Sorting out the FACS: A devil in the details. A very recent Cell Reports paper (from two weeks ago) in which two labs collaborate to find out why they can’t reproduce each others’ FACS experiments.

Characterization of a SAM-dependent fluorinase from a latent biosynthetic pathway for fluoroacetate and 4-fluorothreonine formation in Nocardia brasiliensis. We recently published a paper in F1000Research that shows the same results as a paper published by another group earlier this year, thus independently confirming the validity of the results.