Imaginative Fields in protein folding

Here’s a rough abstract for a talk I’m planning to give next spring, plus a back-and-forth discussion between me and Lily H, a friend and colleague at the University of Chicago.


Scientists use landscape visualizations in their study of protein folding as a cognitive tool to imagine new explanations. The landscape metaphor represents the way that a protein folds into its biologically active state as a traversal across a landscape of possibilities. Each shape corresponds to an energy value, and physical forces drive the protein from higher energies (mountains in the landscape) to lower energies (valleys). The landscape metaphor facilitates scientists’ cognitive work by allowing them to relate the many-dimensional process of protein folding to a physical, three-dimensional shape. For instance, different overall topographies of the landscape stand for different scientific theories of how proteins fold. Scientists have also used landscapes to represent empirical measurements of proteins’ behavior in the lab. I argue that these visual representations of natural phenomena link scientists’ abstract conceptual reasoning to their embodied experience by establishing a spatial field for imagination. I also suggest that this role for images in science — as an “imaginative field” — is quite general, and can be found in many places, such as with phase spaces in physics or evolutionary trees in biology.

Will you talk about particular groups/investigations?

–I have in mind a couple major competing theories for protein folding. The most famous is that proteins don’t follow any distinctive pathway to reach their active state but have to pass over a uniform energetic hump involving internal reconfiguration after an initial collapse. Later modifications to this theory allow for certain biased routes analogous to passes in a mountainside. There’s another case of reasoning with landscapes that turned out to be wrong but is particularly clear in setting out the reasoning process.

Will you talk about the historical use of topographical representation-as-explanation in biology?

–I can, but this links up to the larger topic of fitness landscapes in evolutionary biology, which are conceptually distinct from energy landscapes in protein folding. My feeling is that 25 minutes [for the talk] won’t even be enough to do energy landscapes justice…

Just from the abstract it strikes me that your conclusion could go further—for instance, not just a spatial field for the imagination but what kind of spatial field?

–Maybe, but I’m not sure what kind it would be! Any suggestions? 🙂 This is a place where I could probably learn a lot from existing literature and group feedback. The first idea that occurs to me is to differentiate kinds of fields based on dimensionality and the mode of embodied interaction. Landscapes for example depend on the metaphors of traversal, pathways, and topography. Evolutionary trees, however, involve branching and the idea of distance (which may be temporal or also embedded in a morphological space). How might you distinguish between fields?

–And could this working image have been otherwise?

Depends on what “otherwise” means. 😛 There are considerable variations in the depiction of folding landscapes, much like what David Kaiser describes for Feynmann diagrams in particle physics. There are also other ways of imagining mechanisms for protein folding, such as a divide and conquer process by which parts of the protein fold locally and then combine modularly to form the whole. My impression is that there is no settled, general answer for whether one is “right” overall. A different answer to that question might point to the recent discovery of chaperones that assist protein folding and the historical assumption that landscapes always have a unique lowest point independent of environmental context. (In other words, protein folding involved the motion toward a global minimum specified intrinsically by the amino acid sequence alone.) The idea of a global minimum corresponding to the active state was historically central to the original popularity of the landscape visualization. I’m not sure what would happen if physicists decided the global minimum was no longer explanatorily important.
–What does this particular visualization explain better than others, and what does it not explain?

The main alternative visualization for protein folding is to draw the atoms in three dimensions and watch them move around. This keeps the physical description of the protein’s state concrete, but makes it harder to identify general physical processes. The hydrophobic forces that cause many proteins to collapse, for example, do not have a distinctive signature at the atomic level other than the protein getting more dense, which can also be caused by other forces. The landscape formulation is aimed at describing generic features to all protein folding rather than the particular motions of individual amino acid sequences.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s