Back in December mSphere, an open-access journal published by the American Society for Microbiology (ASM), announced that they would be trying a bold experiment: They would be allowing authors to essentially serve as editor of their own manuscripts. This experiment has been called mSphereDirect. A lot of emphasis has been placed on trying to speed along the editorial process or how people might game the process. Missed in the conversation is the ability to select reviewers that have credibility in your subdomain of microbiology that will help to improve your science rather than stall it.

The concept behind mSphereDirect is that you can solicit reviewers to provide a review of your manuscript. The requirements for who can serve as a reviewer are quite stringent. As an editor for Applied and Environmental Microbiology I can tell you that I regularly use reviewers that do not meet the requirements for the number of publications or level of training that is expected for mSphereDirect. The reviewers also have to agree to have their names openly listed as the reviewers of the manuscript. Once the reviews come back, you then incorporate the reviews into the manuscript, prepare a response to the reviewers, and then get approval from the reviewers that you have been responsive to their reviews. You then submit the manuscript to mSphere with the review history and the names of the reviewers. Within 5 days the editors at mSphere will render a decision and it enters the rather quick publication pipeline.

At the outset, I should reveal that I have several conflicts of interest here. This summer I will be taking over as Chair of the ASM Journals Board ^*. Also, the Editor-in-Chief of mSphere is Mike Imperiale, a colleague of mine in the Department of Microbiology & Immunology at the University of Michigan. But more than that, I’m a scientist that depends on scientific publishers to help me get my trainees’ and my work out to the public. When I first learned of mSphereDirect, I had some of the same concerns I had about PNAS’s and mBio’s fast track processes for members of the National Academy of Science and the American Academy of Microbiology, respectively. mSphereDirect has a lot of overlap with those processes. I think Mike and others have learned a lot from those experiences and have worked hard to put in safeguards to make sure we don’t see those problems repeat.

One of my goals as Chair of the Journals Board will be to bring a more empirical approach to scientific publishing to make it more … scientific in approach. So I look at mSphereDirect as an experiment. Mike and others will collect data, use that data to test the hypothesis that authors want greater control of the editorial process without compromising quality, and then he’ll form a new hypothesis, and collect additional data. I appreciate people’s concerns about putting authors in control and what they would have done if put in charge - those concerns will likely help us concoct the next experiment. I would also remind people of how much we love it when a guest comes to our lab meeting or a third reviewer comes along and only throws stones at an experiment rather than waiting to see what happens or providing constructive feedback. No one designs the perfect experiment. People need to be encouraged to experiment so that innovation can happen. Scientific publishing desperately needs innovation.

In the roll out of mSphereDirect there has been a lot of emphasis put on speeding the review process. I worry that lost in the marketing is what I find most appealing about mSphereDirect. The process goes from being advesarial to collaborative. The dreaded Third Reviewer has become an academic internet meme. You know - the reviewer that doesn’t have a clue what you did but would have done it their way instead. All of us have gotten reviews back from someone where the subject material was clearly outside of their expertise. Worse, we may get back a summary statement from the editor, where they failed to synthesize, highlight, or diminish the reviews. With mSphereDirect, you select those experts that you think would have the most to offer regarding your work. Then you can have a conversation with the reviewers to synthesize, highlight, or diminish their reviews. Taken with the right attitude, the review process goes from being adversarial to working to improve the quality of the manuscript. If this is achieved, then I would call mSphereDirect a success.

I am approaching mSphereDirect as an experiment and based on my N of 1, I think the ability to cutout the Third Reviewer will be more significant than the issue of speed. In December we posted a pre-print for our latest manuscript to biorxiv. I then sent a copy of the manuscript and instructions to three people with bioinformaticists expertise that were at my academic rank or higher and who I thought would do a good job of evaluating our work. Naturally, two declined the invitation and one accepted. I invited another individual and, thankfully, they agreed. Naturally, being around the holiday season they requested a delay in the deadline. So far the process is eerily similar to what I do at AEM as an editor! From inviting reviewers to getting a final decision, it will likely take just as long as the traditional route. But… I get to pick the people that I think have the closest expertise to evaluate what I am reporting in the manuscript and I hope that depending on their comments, we can have a discussion about how best to improve the manuscript prior to submitting it to mSphere.

Regardless of the outcome of this experiment, I applaud ASM and mSphere for taking this risk to innovate improvements to how peer review is done. It might be a flop - many experiments are - but we will learn something and be in a better position to launch our next experiment.

* It should be painfully obvious that I don’t speak for ASM, the University of Michigan, or anyone by myself. ↩