Disclaimer: I am an ambassador of the protocol sharing website protocols.io. Yes I got an $50 amazon voucher from them end of last year with which I bought some books for my son Saul. I think pre-prints are awesome but have some caveats. And foremost I love free stuff accessible by everyone.
I am currently working on long-read genome assemblies of fungi and plants. All successful long read genome projects start with months of optimization for DNA extraction methods. This can often involve phenol chloroform extraction for weeks on end with slight alterations of the same protocols. Mine took about 9+ months to optimize and to finally get some good data back. I was lucky as I got three protocols to start with form colleagues around the globe. Looking into the scientific literature was no help at all. The one I build on was from INRA and put online by Jason Stajich here, which I only found because I asked Andrii Gryganskyi for help. So overall I didn’t reinvent the wheel but simply build on what others already created previously. Pretty much was 99%+ of all science really is. Notwithstanding, I knew several other groups struggled with the same problem so I put my working protocol online on protocols.io and publicized it on the 18th of April 2016.
— BenjaminSchwessinger (@schwessinger) April 18, 2016
Now on the 31st of August 2017 we published pretty much the same protocol as a peer reviewed methods chapter in a book a friend of mine edited.
— Philipp Bayer (@PhilippBayer) September 7, 2017
I would like to celebrate this occasion by sharing some thoughts about open sharing of methods early. Overall I obviously think it is advantages for you and everyone else in science to do so. This is in contrast of letting one’s protocols rot on one’s harddisk and to only let them out to see the light of the day once someone offers you a co-authorship.
In semi random order.
The most obvious one first. It took 17 months to publish pretty much the exact same protocol I knew was working when I put it up online. That’s nearly half my son’s life. In the meantime the protocol got 2375 views, I got 10+ emails or messaging requests to explain different parts of the protocol better and some ‘Thank yous’ on conference as other got it to work for their favorite species as well. Overall putting it up online early enabled other to discover more in in less time instead of reinventing the wheel of DNA extraction protocols.
I know some people would not use none peer reviewed protocols that they find online (discussion here and here). I personally think it is an illusion to believe that methods get properly reviewed in papers. I only ONCE got a comment on a small technical detail in my methods section across all my 34 publications. Similarly, I’ve hardly ever (never?) seen another reviewer ask for details or even comment on the methods section.
I know the protocol worked for me and no closest peer review would have changed it.
Discoverability and ease of use
Everyone loves detailed easily accessible protocols. EVERYONE! Everyone despises short methods section in the supplement that provide no detail and cite another paper that cites another paper that cites another paper that cites another paper that supposedly did the experiment the exact way the authors did it.
few things frustrate me more than having to go 5 or 6 references deep to find details on the methods in a paper 1/n
— Michael Eisen (@mbeisen) September 9, 2017
So putting your full detailed protocol on your lab webpage is good (see this awesome long read DNA extraction protocol) yet putting it up on a publicly accessible and search-able platform like protocols.io is even better. You get a DOI, it has versioning, people can fork it, people can comment on it, people can ask you directly, it is backed up available forever, and you can even add it to collections/usergroups so even more people can discover it. Everyone is happy using detailed protocols they can interact with and cite. Everyone!
Catering to both systems
Of course some people get nervous and fear they cannot publish their methods in esteemed journals anymore. Of course not everyone goes online on novel platforms like protocols.io to hunt down the latest protocol. Many prefer to search the peer reviewed literature. And that is really the great thing about depositing full methods early. You can do both. You can share your protocol to help others quickly and get the bonus points of a journal publication. That is just what it did! It might well be that more people will cite one’s publication as they already used one’s protocol before it was officially published in a peer reviewed journal. That is the case for preprints and future studies will show if we can observe similar effects for protocols protocols.io. For sure I will cite both protocols.io and my book chapter in my next genome manuscript preprint.
Reproducibility and rigor
There is a lot of talk about the lack of reproducibility in life sciences. Much of this is focused on depositing code online. Not much is focused on the actual data generation. Methods section are often poor and most supervisors don’t care a la ‘just write something no one reads it anyway’. Yet every good project starts with good data, starts with a solid detailed protocol of how things are done and were done. This not only helps others to reproduce one’s work but also oneself and one’s lab. Writing down great protocols often makes one realize the fine grained and important details. Sharing those protocols publicly is even better as it accelerates discoveries, saves money, makes science a better place for everyone. On top we in the western world sitting at elite universities often forget how hard it is to access high quality protocols. One just need to look at researchgate to see what I mean.
Come join us, make everyone happy, and share your detailed protocols online early. Imagine a world were you just have to search for a detailed protocol and you find one pretty much immediately. Just like cooking recipes.