Gaurav Vaidya

Topic pages and biodiversity informatics

PLoS Computational Biology has embarked on a fascinating experiment in scientific publication to Wikipedia: Topic pages. That's a peer-reviewed article published in PLoS Computational Biology designed to be directly incorporated into Wikipedia. Unlike a conventional research paper, a topic paper is designed to comprehensively cover one specific topic. The first ever topic paper was about Circular permutation in proteins, a Wikipedia article which was significantly improved by the changes.

In one elegant step, this ties together Wikipedia and scientific credit: once your topic paper is published, you get the credit due to you, while Wikipedia is greatly improved by your contribution. It also ties together Wikipedia and journal publishing: one of the big problems in taxonomic informatics is tracking the bleeding edge as new species are discovered and named. By writing the articles in a wiki before submitting it for publication, the process of going from scientific paper to Wikipedia article can be simplified and sped up. Finally, by framing the goal of the paper as a "topic paper" rather than as a research question, the content is already in an encyclopedic tone and subject incompatible with a typical, hypothesis-based research paper.

Will we perhaps someday see taxonomic topic pages in which the cutting edge hypotheses about a clade's phylogenetic tree are presented, ready for inclusion into an encyclopedia? Or perhaps taxonomic review topic papers, with the most comprehensive knowledge assembled about a species or clade, summarized by experts, ready to be incorporated into the most widely used encyclopedia on the internet? I do hope so. Daniel Mietchen, the Wikimedian in Residence on Open Science, is already interested in taxonomic treatments, and I can't wait to see what he -- and the PLoS and OKFN teams -- come up with next.

The #SciFund Challenge

I just discovered the #SciFund challenge, a Kickstarter-like project to raise money for scientific projects by promising deliverables created as part of the project. I think it's a great way of getting non-scientists involved in science, to say nothing of getting funded and of creating a concrete science-to-public deliverable system. So: excited!

Here's some of the projects I took a fancy to. Unfortunately for me, as a grad student lacking money, RocketHub (the hosting website) has no mechanism for "subscribing" to a project without paying that I can find. So this might be the only way for me to keep an eye on them (and maybe make a contribution with my next paycheck).

• Collecting ants living nears homes and schools, which sounds like an awesome way to get kids involved in science.
• Genetic surveys of Costa Rican birds and bats, the American pika, and the social tuco-tuco. Because Ctenomys is to love me.
• Hiring local field assistants for elephant research in Sri Lanka, building closer links between researchers and the human residents of the landscapes they study.
• Using Ancient Rome-era DNA to determine the geographical origins of immigrants to the Empire.
• An infrastructure project to deliver vaccines to remote areas by matching up travelers in rural Tanzania with vaccines which need to be delivered to particular villages. That's a really smart idea!
• Do camera traps affect the behavior of their study animals?
• Why do some spinner dolphins have "reversed" fins, which is a fascinating study, mixing up biophysics and evolution/selection pressure research in sexy, sexy ways.
• Tracking dolphin populations through photoidentification, whose project description reminds me that labradors can track killer whales. Which is an awesome thing to be reminded of.
• Parasitic trematodes which change the behavior of the Californian killifish so they get eaten by birds (where the trematode can finish its life cycle).
• Evolution between sunflowers and bees in agricultural fields.
• Explosive duck penis erections, which really needs no elaboration.

The most fascinating part of this enterprise for me is the connection between the science projects and the contributors: the deliverables. Rewards for donating ranged from credit (on websites or in publications), to a copy of the research results, to exclusive access to project blog and Twitter feeds (which make no sense to me: wouldn't you want those feeds to be as public as possible, to encourage more people to join in and contribute?). More interesting deliverables were also on offer, from elephant dung paper and calendars with photos of the research subject, to photos taken during the study, clay sculptures of dolphin fins, lyrics to a Tanzanian children's song, and comics personally drawn for you by Zach Weinersmith of Saturday Morning Breakfast Cereal fame. Allegedly, a comic of "Carl Sagan riding a unicorn" is not out of the question.

One thing I did find odd was the lack of links in the project descriptions. Only a few projects linked to websites, blogs, Facebook pages or Twitter feeds (one project going so far as making the Twitter and blog feeds "contributor only"), although just about all of them had really excellent websites hidden away in the "About Me" section. I love Twitter for letting me follow a project without getting too involved, so more Twitter feeds would have been great to see.

Thinking in the longer term, I wonder how well this would work without the #SciFund banner, though. Unlike music or movie crowdfunding, where artists can put up samples beforehand, it might be harder to prove your scientific worth to a broader audience. For instance, how would I differentiate between a serious project to study bee evolution and some nutjob trying to prove that evolution didn't happen? I could look at the project description, sure, but it's not hard to reference many scientific studies to give yourself an air of authenticity. Perhaps the #SciFund banner will eventually be replaced with institutional banners: if the research is being carried out by the Smithsonian, the University of Colorado or is partially sponsored by National Geographic, you can be pretty sure it's been vetted.

Some tips on screencasting on the Mac

We've made a screencast to promote Sequence Matrix, a genetic sequence concatenation tool I wrote at the Evolutionary Biology lab at the National University of Singapore. This was my first attempt at creating a screencast, so I learned a thing or two I'd like to remember the next time I make one of these.

This video was recorded and edited on a MacBook Pro using ScreenFlow, a piece of screencasting software I chose after surveying screencast products late last year. The audio files were recorded by Prof. Lohman using iMovie and uploaded to a file sharing service. This means that I got them as individual clips, which were much easier to work with instead of a single, large file which I'd have to cut up myself.

Here's what I'll be keeping in mind the next time I need to make a screencast:

• Use a separate account with minimal decoration: create a new user account on your Mac just for recording screencasts. In this user account, get rid of the status decoration on the status bar in the top-right corner of your Mac screen (you can drag them off by ⌘-clicking on them). Reduce your screen resolution: 1024x640 worked well for me. I think YouTube uses a 16:9 ratio on its videos, so this screen ratio (at 16:10) will be stretched slightly, but it won't be too bad. This also means you're using your entire screen, which really shows off the Mac's gorgeous user interface.
• Get rid of the time: beware of the clock in the status bar (upper left corner) and iCal (on your Dock), which will show the effect of your editing by jumping forward and backward in time. Even worse: record on multiple days, and both these apps will be changing days as well. Longtime Mac users might not realise that recent versions of iCal continue to show the correct date even when not running - the days of iCal Day has, it seems, come to pass.
• Pick a homing point to move your mouse pointer to in between clips. For this video, my homing point was about 2/3rd of the way down from the top of the screen and 2/3rd of the way from the left of the screen. I'd start recording a video clip with my mouse pointer there, and end the clip by moving the mouse pointer back there. That way, the mouse doesn't appear to jump around in between spliced cuts. Another added benefit - if you later decide to rearrange your clips, each video will continue to start and end with the mouse at the same place, so they'll all continue to fit neatly together.
• Use speedups to demonstrate something boring: I'm not so sure about this one, but it's something to think about. The first time we export a file in the video, I sped it up slightly (150% or so faster than usual) so as not to bore the viewers. The second time we export a file, I run it at between 300% to 500% normal speed. That way, the viewers can see what's happening (and realise that yes, it's pretty much exactly what they saw earlier), while also moving swiftly along to the interesting bits. I did this a lot towards the end of the video (during the Nexus and TNT exports), and I'd love to look at it again with fresh eyes in a while and see how it came out.

I'll move these tips into the Creating Screencasts WikiBook when I get the time, but for now please feel free to add your comments, suggestions and complaints on the YouTube video itself or on this blog post.

The joy of experimentation

It's far too easy to forget just how wonderful experiments are. It's a scientific paper in a nutshell: you show what you've got (Materials), you do something with it (Methods), something happens (Results) and then you get to discuss it. And not just "something happens", neither: so many of the best known experiments have astonishing and eye-catching results, from dramatic colour changes to peculiar vibrations to the completely unexpected.

A few weeks ago Karen James linked to Sixty Symbols on her Twitter feed, introducing me to the works of Brady Haran. Brady records scientists talking about physics (Sixty Symbols) and chemistry (Periodic Videos) at the University of Nottingham, edits them into tight little bundles of experiment, anecdotes and discussion, and puts them up on YouTube. I don't know how effective these videos are at connecting to non-scientists, but - as a long time science enthusiast - they're right up there with the best science shows I've seen.

Each short video focuses on a particular topic: carbon, angular momentum, or Planck's constant, and consists of discussions, mini-lectures and experiments - lots and lots of experiments.

Not that these videos are just for science newbies: Periodic Videos can go to places your high-school chemistry teacher simply couldn't. You can watch flourine being synthesized (and then used, MythBusters style, to set many things on fire) or aqua regia dissolving gold. And when you watch a video of a physicist waxing lyrical about his favourite number, you know you're in safe, geeky hands.

If you're at all into science, you really need to check out Sixty Symbols and Periodic Videos.