Biocurious is a weblog about biology, quantified.

Cell Visualizations

by Andre on 2 October 2006

A collaborator just sent me a link to a beautiful animation showing some of the inner workings of a cell. Sean at Cosmic Variance also made reference to it in a recent post. You can see some other work from the same company at their website. The video was made in collaboration biologists from Harvard, who ensured that the structures depicted in the animation were accurate, but it fails to convey the crowdedness and chaos that characterize the cellular environment. Small is different, and not understanding that can make Intelligent Designers’ arguments about complexity a lot more appealing. After all, how could proteins seek each other out and bind in the right orientation in the absence of thermal motion? How could those actin filaments assemble themselves like that if there’s nothing pushing the parts into place? Being good PR types, they know this and even produced their own video depicting transcription and translation that is also accurate in its own way but is specifically designed to sound implausible (the cellular part of the movie starts about a third of the way in).

But it doesn’t have to be like this. Given the talent of the scientists and animators involved, I think the project could have been more accurate with much more educational value if they had gone to greater pains to incorporate more than just accurate structural information. For example, crowdedness and clarity could both exist happily in the same sequence by taking sections, using different colours for different parts, and making creative use of transparency. Components that are initially opaque could accurately reflect the number of cellular components and then selectively dissolve and reappear to highlight different features of interest. A similar technique could be used for fluctuations. Temporarily damp and/or slow a component’s fluctuations to demonstrate some aspect of its function like the walking motor tugging its cargo pictured above. Students will develop their intuitions based on these movies and the related diagrams they see in their textbooks and could easily come away with erroneous ideas about cells and what is possible in the cellular environment.

An even better way to convey these concepts would be through an interative presentation. That might sound like quite a challenge, but it’s not impossible. If you’ve ever experienced a modern video game you’ll see how successful today’s developers are at creating immersive interactive environments. Just give these people some accurate structures and watch the magic happen. This isn’t my idea though, and the approach already has a name—Data Gaming—and it’s currently being developed by a company here in Philadelphia called Tacitus. They don’t have much on their website yet, but I hope that will change in the future. I’m looking forward to the first biological applications of their concept.

  1. PhilipJ    3760 days ago    #

    I wonder if the real issue is computers. I assume making these videos is roughly equivalent to MD simulations that people are doing as serious research, in which case the computer {power, time} needed for even a nanosecond worth of brownian motion of a few hundred atoms is huge, and given the wide variety of cell locales they depict, how long it would take to get reasonable movies.

    I agree, however, that my first thought when watching the movie (after “neat!”) was, why are structures the only aspect which was required to be realistic? The physical characteristics of cellular interiors are too often ignored.

  2. Fred Ross    3760 days ago    #

    I think most biologists, though they might make some noise about Brownian motion, at a deep level believe in little cartoons (or in more recent years, static structures) flying around in simple patterns. It’s certainly the pattern of thought the textbooks promote. Is it surprising, then, that a movie instigated by biologists looks like that as well?

  3. Andre    3758 days ago    #


    I don’t think it needs to be as computationally intensive as MD since you’re not interested in such small scales in space or time or in rigorously correct results. They should just look correct. Pseudo random motion of particles would probably be fine. I did a search on youtube for video game physics and came up with this:

    Seems pretty good to me. Less particles, I know, but things develop quickly.


    I think it varies widely from person to person and who you include in the definition of “biologist,” but you could be right. It’s interesting that you bring up text book figures because I’ve heard videos related to the one that we just saw are supposed to form the centerpiece of a new Harvard undergrad curriculum in cell biology. Same structures and concepts, now with inaccurate motion! It was just a demo though and I do think animations can help convey more information and be more memorable than figures.

  Textile help