Review of Open Science Summit 2010, #OSS2010

I have been attending Open Science Summit 2010 at Berkeley, CA and although not quite finished yet I feel like I can give an overall review of what I thought of the conference. You can check out my individual comments during the conference on Twitter.

I would like to state that in general I am grateful and respect the work that Joseph Jackson and the organizing committee conducted to make this open science conference a reality. It is a tremendous amount of effort and the following is only meant as a constructive criticism for possible open science summit conferences in the future.

Pros

• Bringing together a very intelligent diverse group of speakers. Good mix of policy makers, developers, traditional scientists, biotech, young and old, etc.
  
• Great use of technology. Providing a live video stream of conferences is an idea that I wish more conferences implemented. Also, using backchan.nl is a nice additional add-on that couples well with the live video stream.
• Willingness to try to adapt (as much as possible) to conference attendees comments via twitter, back channel, etc.

Cons

• No scheduled breaks. Breaks are needed for numerous practical reasons: people need bathroom breaks, time to get some fresh air, and time for talks to get back on schedule. Even more importantly, it allows people to mingle. People travel to conferences so that they can get a chance to connect with people face to face (otherwise they would just watch the online feed).
  
• No time for Q & A. Questions immediately after speakers not only is informative, but gives a temporary "mind break" for the audience. It also gives time for IT to get the next presentation queued. Note: this did tend to improve as the conference proceeded.
  
• Too many speakers. Having 25 speakers in a single day (without parallel sessions) is just too much information for people to take in and sit through.

Additional lessons learned

• A no slide presentation is not a guarantee that it will be a good one.
• Videos do not always make a presentation better.
• Having 2 or more speakers from the same organization or having the exact same opinion is not really beneficial.

Comments from FriendFeed

Review of Arrowhead Conference

I was hoping to write this during the conference, but every talk I went to was really interesting and I ended up not using that time for blogging. However, there is at least one talk that I wanted to mention before I forgot to much.

I thought the most interesting talk was from Kim Lewis, who describes "persistor cells" as those bacteria that lay dormant in a population. These persistors are resistant to antibiotics because they are essentially shut down and passivly allow antibiotics to simply wash over them as opposed to normal bacteria cells that actively block or pump out antibiotics. Lewis also showed that late samples taken from cystic fibrosis patients had high levels of persistor cells. Kim then discussed unculturable bacteria (of which 99% of bacteria are) and suggested that when plated on media that these are actually dormant and not dead. To support this Kim showed that by innoculating an unculturable sample with E.coli caused growth of an unculturable strain around the E.coli spot. He later found a mutant that did not cause the effect and identified the key gene to be a sideophore. Lewis ends with this little tidbit, "Dormancy is the default mode of bacterial life". I find this really interesting because it suggests that most bacteria depend on a few bacteria to signal when their surrondings are optimal for growth.

There are a couple of more talks that I hope to blog about in the next day or two.

Microbial Genomics

I have been at Lake Arrowhead since Sunday for the 16th International Microbial Genomes Conference and I have to admit I am quite impressed. I think this conference has solidified in my mind that large conferences can't compete with smaller conferences. Let me list the reasons why in order of importance:

1. Food - in almost all cases the larger the group of people the worse the food will be. Now you may think I am slightly joking around saying that this is the most important, but I am quite serious. There is nothing worse than having to eat some cafeteria style food and then have to sit through 2-4 hours of talks with a cramping/rumbling/starving belly. Also, I find meals are the best place to meet and have discussions with other scientists.
   
2. Meeting people. You get a chance to meet almost everyone you want to without having to hunt them down like a gazelle. I really detest pouncing on a speaker as soon as they are done a talk. It is much nicer to see them at a break or at a meal (see above) and introduce yourself and ask a question then.
3. Better science. I find at smaller conferences the talks have been hand selected and tend to have a better line up of speakers
   
4. Location, location, location. Smaller conferences tend to have their meetings at nicer locations.
5. Beer & Wine - From my experience alcohol tends to be cheaper (or free) at smaller conferences which always makes everyone happy and tends to get scientists to loosen up some.
   

Open Source Conferences

Last Thursday I found out via Jonathan Eisen's blog that there was a live stream of the Metagenomics 2007 conference. I tuned in over the next couple of days and got a chance to listen to some great presentations. Although, I didn't have any questions I had a couple of offers from people I knew attending that were willing to ask for me. How cool is that?!

I hope that more conferences start webcasting proceedings and/or making the recorded presentations freely available on the internet.

Regardless, there are a few things that are necessary to make a good webcast:
1) All presenters AND questioners must use microphones (this is a good idea regardless of webcasting)
2) High resolution video so that details of the presentation slides can be seen (this is usually a major problem)
3) A nice cross platform streaming media (no windows streaming media)
4) Good encoding so that the majority of internet users have enough bandwidth