Tracking Words and Fire: The vast geography of activity maps

These past two weeks I have been on a ‘data expedition’, sorting through about 3.4 million data points (14 data sets, each containing ~50 neurons, over 1200 time points for four conditions). I admit that I do not have a lot of enthusiasm for my data mining; it has been a slow, painful process. The goal is exciting though – I am seeking emergent characteristics in neurons that recover from an injury, distinguishing them from those that perish. In particular, do certain neural activity patterns promote rapid recovery from brain injury? 

Well, stay tuned…I have a lot of staring at the computer to do. Along the way, I find myself paying attention to ‘activity maps’ in general – and wow, there are some fascinating-odd-big things out there. 

Here is a small curation….

First, we have the 3 billion dollar one: Brain Activity Map (BAM).

Definitely read the short paper linked above for a good explanation of the goals. The project plans for a concerted effort to develop new tools to precisely measure brain activity, starting with simple organisms like C. elegans (a transparent roundworm) and moving up the metazoan ranks to fruit flies and then possibly the smallest mammal, the Etruscan shrew (just a few million neurons…). Currently, the best data we have on brain activity come from calcium imaging (thank you dear jellyfish!). The technology for calcium imaging is pretty fantastic. 

Nature Methods (2013) Doi:10.1038/Nmeth.2434

Here are the latest: from a larval fish brain (Ahrens and Keller, Nature Methods, 2013), but the scientists of the BAM project want to attain higher temporal resolution by measuring voltage changes, which cause the changes in neuronal calcium. There are a number of naysayers who claim the proposal to be a political stunt with poorly defined goals, but I think in the hands of the right scientists, the emergent activity maps will be well worth the funding.

Meanwhile, in the world of the social sciences, the data are abundantly popping out of smartphones with more than adequate temporal details. Social networking data have been making some very odd maps, including one from a ‘hedonometer’ that predicts the happiest and saddest cities in the US by measuring the frequency of ‘happy’ and ‘sad’ words from Twitter data: hedonometrics.

It appears that the abundant use of the word ‘sh*t’ in the tweets of Beaumont, Texas has made it into a ‘sad’ place – happier than Louisiana though, possibly from their abundant use of lol. Meanwhile, the inhabitants of Napa California maintain contentment with beauty, Christmas and of course, the keyword: wine. By the way, guess what one of the happiest words is? Yeah, rainbow. 

Here is a very dynamic social networking map. It is from a smartphone app called foursquare, showing check-ins to restaurants, clubs, etc. in two cities as dawn progresses to twilight. Watch the change in colors indicating the type of activity over time. 

The new social data are quite cool, but I found my absolute favorite activity map in an old film, For All Mankind. This is a beautifully composed documentary on the Apollo missions of 1968-1972. At about 14:32, the astronaut is looking back at the earth as they gain elevation over Africa. Down below, he sees small points of light dotted throughout the dark, open plains. He realizes that they are campfires from nomads resting for the night. Maybe dots on a plane are not as visually alluring as electric movement through a city, but I love that the context is clear and pure, making it the easiest to connect to.

If you don't mind a foreign link, here is the full-length film: http://www.tudou.com/programs/view/8DvtsjYg53g/

Okay, I think I am inspired enough now to return to my maps.