Monday, January 19, 2009

Writing our first paper--Shotgun DNA Mapping

Today, our lab created a sort-of-complete draft of the first paper to come out of our lab. I wrote briefly about it on our kochlab research blog and also posted our draft on scribd. I don't want to re-blog everything here so I'll just say that we're really excited about the paper and if you happen to click on those links above, I (we) would love to hear your opinions.

I'm really proud of the lab members, and especially Larry, who did most of the programming and produced the results that are in this paper. He produced much of this during two amazingly productive months last year (say August/September-ish) when he was seemingly-happily working 7 days a week with many 10+ hour days. If you happened to read my previous blog about talents in the lab on my science blog, you may have just realized that that kind of dedication and productivity is indicative of some kind of underlying talent that Larry has, and I wholeheartedly agree. Larry and I recognized this during our recent "talents meeting," and we're wanting to analyze that time period as a way of more clearly seeing what these strong talents are.

In contrast to the research productivity, we've been wanting to draft this paper for several months now, and up until today it was painful for both of us. I definitely committed some management sins (I'm avoiding a euphemism we use in this situation), but I think he's forgiven me by now. Last week, we fully consciously realized that the writing process we'd been attempting was failing and so we decided to try something new. The process we had been trying was to collaborate on our private wiki by having Larry draft sections and I and others commenting on his work and making suggestions. There are probably many reasons this failed and I know many of them are due to non-talents I have with writing and management. The new process we tried today was to just block off a whole day in a conference room to work together towards a single outcome: to produce a document in Word that was a reasonably coherent draft of the paper that we could distribute to our friends and collaborators for advice. Anthony, Linh, Larry, and I did this today, and I think the experiment was highly successful. Certainly the draft we produced is very rough, and none of us will be surprised if >75% of it is changed as we revise and pound on it in the coming weeks. But I think we surpassed the outcome we were looking for and for sure we have achieved a very solid foothold towards our goal of actually publishing this paper in a refereed journal.

This process took us about 9 hours together, plus who knows how many hundreds of hours of discussions and prior writing attempts that we built upon. We were able to fight through several instances of writer's block today, particularly with the introduction, and I think the clear outcome we were pursuing was a huge help in pushing us through this. Many may question whether it's practical for me (or any of us) to schedule an entire day for writing a paper draft, and I hold the same question. I'm not sure whether we can or will need to do this in the future, but given the success, I think we'll try. Time like this is precious for all of us, but we received quite a valuable payout. I present my first lecture tomorrow (in about 13 hours) and I haven't made any preparation except for what I have from last year and a bunch of mostly subconscious scheming and worrying. I'll talk about this on my teaching blog, perhaps later tonight. I'm super-happy about the progress we made today, and very proud of my students ... and I also now have a really uneasy feeling in my stomach in anticipation of being unprepared for the onslaught of teaching and the grant deadlines. I should mention that the uneasiness is probably compounded by hunger, which would be the subject of another blog that I don't have.


  1. 9 hours is a very good benchmark for a paper like this!

  2. One of the interesting things about collaborative writing processes that I am finding is the absolute necessity of a strong and coherent outline. This tends to mean that someone has to take the lead or you have to take the "lock em all" in the room approach until you get something like that together. I would say nine hours sounds about right to me.

  3. This reminds me of a recent style of agile programming called a "sprint".

    I find that while I have written dozens of pages into wikis, I have never written an essay or article in that format. I speculate that wikis tacitly encourage linking --- and each switch of a web page to verify a link causes an interruption in thought. We feel like we are being productive, but actually, we are interrupting ourselves.

    Steven Jenkins, "Concerning Interruptions", < >

    Joel Spolsky, "Human Task-Switching Considered Harmful" < >

    I find wikis great for storing tidbits of reference information, gradually accumulated over time, but for writing a stand-alone document, they force you first to decide, "where do I want the document to go" before "what do I want the document to say"; and for scientific writing, I propose the latter is more important.

    A long time ago, I also had a problem where the wiki would log me out while writing a long document, which caused me to lose work. This doesn't happen with Microsoft Word.

    I am curious about the logistics. Did you have one computer or two or four? LCD projector? One typist or two? Did you have the plots and graphics beforehand or did you create them on the fly? How many breaks did you take? Did you break for lunch or order food, and if the latter, did productivity suffer after the food arrived?

  4. Richard, I was going to comment earlier, but got distracted by your links. Ha ha, just kidding, I haven't actually looked at them yet. Do the references (or worse footnotes) in published papers do the same thing? Probably not as bad...

    Thank you for your comments and really great suggestions on the other pages. We're going to try to respond within a couple days.

    That logging out problem with the wiki was aweful -- it really burned some of my Junior Lab students bad. Sorry you lost information, that's really bad.

    As for logistics, we started at about 9:30 am. We sent Linh to McD's for lunch--dollar menu items for those not on a diet, and SW grilled chicken salads for Ant and I (we're on a weight loss competition). We probably had lunch for about 1/2 hour, not talking about the paper. I and Linh brought chips and other stuff for snacks--I would have eaten 4,000 calories of corn chips if not for the competition. Next time, I will bring a whole lot more diet caffeine pop, and try also to have better snacks and beverages. We had one projector, and only one computer that could type. Everyone had a computer to either look stuff up, or work on other stuff. I think it would have been better with VNC so everyone could have immediately taken control of the typing. We switched off typists every once in a while. The white board was an essential part of it. We only used graphs that we had created earlier for other purposes (talks, etc.). I'd say we only had 1 to 1.5 hours of breaks, and we were working very well most of the time.

  5. This comment has been removed by a blog administrator.


Creative Commons License
This work is licensed under a Creative Commons Attribution-Share Alike 3.0 Unported License.