Jun 202011

I hope I’m going to submit my PhD student’s first comparative genomics paper very soon. Three of us have written the manuscript collaboratively using Google Docs. GDocs is an online word processor and although I’ve used it quite a bit before, this is the first time I’ve used it to write a manuscript with colleagues. Its been (almost) excellent, here’s the review.

I’ve had a number of manuscript experiences where I’ve spent long hours trying to collate different authors’ contributions into the same Word document. The idea of using GDocs is that multiple authors can have the same document open at the same time making changes without any conflicts or the whole thing crashing. You never have to ask “which is the live copy?” since there is only one copy.

Good parts

  • Nobody has had to collate mutually incompatible versions into one document and circulate (again) for people to check.
  • It is a clean GUI and a pleasure to use.
  • There is a good comment system and these are supplemented with a realtime discussion panel, just like Skype or other IM client.

Less good parts

  • It doesn’t work well in some older browsers. Tell collaborators to use Chrome, otherwise they may complain that its a bit rubbish and doesn’t work properly. Using Chrome there are few to no problems (ie better than Word).
  • You cannot use any sensible reference software. Mendeley, Zotero or any of the other reference managers you know will not allow you to insert references and format a bibliography the way you would in Word.
  • Track changes is not as good as in Word.

Google Docs revision history

Overall its been great I think. There are a few things I really wish were different. Track changes could be easily improved to identify who has done what. Yes versions of the document can be compared, and rolled back to previous versions, both of which are useful but none of it is quite as obvious and easy to use as in Word. More than anything I really wish that reference management was better. We have been typing in place holders (Smith 2000) and then exporting the document as a Word file and introducing the citations using a reference manager before submission. This sounds bad. Why not just do everything in Word? Well, even today, two of us were making some last minute changes on the Word version, each copy with someone’s initials appended and somebody tomorrow has to reconcile it all.

I think Google Docs if adopted widely would have a great impact on writing multi-authored manuscripts. I don’t think it will be very widely used in science though unless reference managers can integrate with it properly. Despite this I have really enjoyed writing a manuscript with it, and, even though it has to be passed through MS Word at the end, on the whole I’ve much preferred it.

Sep 122008

I just read a post at Mario’s Entangled Bank called “Running an Academic Lab Google Style“. Some interesting ideas that I was aware of but had kind of forgotten. When I was a postdoc I had lots of strong views on how research should be done and although I haven’t really changed my mind it is so much more convenient to not think too much and just do research the way everyone else does.

The thing that got me thinking was the idea that at Google engineers get 20% of their time to work on something company-related that interests them. What would happen if we gave the same freedom to our postdocs and students?

The Google Way: Give Engineers Room (NY TImes)
Google’s “20 percent time” in action (Official Google Blog)
Life at Google

Googling shows that there are a lot of skeptics out there who don’t see how it works or believe that it can work in other environments. Joe Beda works at Google and nicely summarises (Google 20% Time) the reasons why it works for Google and may not in many other environments.

“The intrapersonal environment at Google is very energizing. When someone comes up with a new idea, the most common response is excitement and a brainstorming session. Politics and who owns what area rarely enter into it. I don’t think that I’ve seen anyone really raise their voice and get into a huge knockdown drag out fight since coming to Google.

Can 20% time work at other companies? I’m sure that there are going to be others that try. However, I think that it is important to realize that it is a result of an environment and philosophy to development more than a cause. I don’t think that it is something that can be imposed in an independent way”

I think research labs are one of the few places where this sort of environment would work well and be quite naturally developed.

Funding agencies would probably have a fit at the prospect, although I think the outcome of those grants would probably be more effective in solving the big questions posed in the grant proposal than where the staff followed the initial grant-plan religiously. I honestly don’t believe productivity would be lower, maybe the opposite. Wikipedia reports on Google saying

“In a talk at Stanford University, Marissa Mayer, Google’s Vice President of Search Products and User Experience, stated that her analysis showed that half of the new product launches originated from the 20% time.”

Although I don’t want to lie, I’m putting this on a public blog after all, not explicitly telling the funding agencies would prevent the nice grant-people from getting stressed.

Now I’m not suggesting that postdocs and students be given Friday off to go to the beach. Even Google uses the term “company-related” which I guess would translate into “something relevant to the questions and work of the lab”. It is important to nurture creativity and a feeling of involvement in shaping the research direction and strategy of a lab. Maybe researchers should be actively encouraged to spend 20% of their time

  1. thinking deeply and laterally about big questions in the area
  2. fixing small but annoying technical issues
  3. testing the feasibility of wild and weird ideas
  4. learning about and working on related topics and techniques that may one day feed back into their main project.

I get the impression, though it may be apocryphal, that this sort of thing used to happen much more in the past. Now there are detailed programs of work in grant proposals, few people are on “soft money” and side-projects can be seen as unprofessional. I remember having lots of conversations about side-projects as a post-doc and PhD student. Despite having very supportive supervisors I never really thought they would get it and would see it as a waste of time. I hid those side-projects, discussing with labmates but not the PI, until they were at a developed stage with clear outputs. Looking back those projects have generated my best and most highly-cited papers! The 4 papers listed at the bottom contain my 3 highest cited publications. Think what might have happened if I had discussed other ideas with the knowledgeable PIs of those labs.

PhD; 2/3 papers came from side-projects
Postdoc; 1/3 papers came from side-projects
My first small grant in my own lab; 1/2 papers came from side-projects

Of course my personal research is one big collection of side-projects, I can do whatever I like. But on grant-funded work maybe I haven’t encouraged people to just explore enough. That sounds like a resolution to me- all I need now are the grants! I intend to implement this 20% time with my next appointed student or postdoc, but it may be a while before I know how its working.

The other idea that is strong at Google is talking and mixing with a wide pool of other employees. There seems to be a culture where no ideas are considered stupid. The concepts of interaction and productivity in personal projects are not unlinked. This is something I’ve been really keen on at Hull- Journal clubs, informal lab meetings, communal coffee time, after work pub “meetings”. Most breakthrough science is done over coffee, I am convinced of it, but it is unlikely to be a legitimate consumables expense on grant proposals any time soon.

Some publications from side-projects

  1. Gómez, A., M. Serra, G. R. Carvalho, and D. H. Lunt (2002) EVOLUTION 56:1431-1444. PDF Speciation in ancient cryptic species complexes: Evidence from the molecular phylogeny of Brachionus plicatilis (Rotifera)
  2. Lunt, D. H., and B. C. Hyman (1997) NATURE 387:247-247. PDF Animal mitochondrial DNA recombination
  3. Lunt, D. H., D. X. Zhang, J. M. Szymura, and G. M. Hewitt (1996) INSECT MOLECULAR BIOLOGY 5:153-165. PDF The insect cytochrome oxidase I gene: Evolutionary patterns and conserved primers for phylogenetic studies
  4. Szymura, J. M., D. H. Lunt, and G. M. Hewitt (1996) INSECT MOLECULAR BIOLOGY 5:127-139.The sequence and structure of the meadow grasshopper (Chorthippus parallelus) mitochondrial srRNA, ND2, COI, COII ATPase8 and 9 tRNA genes

Picture; brainstorming at Google, link