Geometry of behavioral states.

<p>This figure shows the two-dimensional arrangement of behavioral states discovered by unbiased open-loop fits. Each circle (except the black one near the bottom of each panel, which represents complete immobility) represents a single state from a single worm. The area of the circle is proportional to the amount of time the worm spent in that state. The gray background in A–F and H, representing all states discovered in all experiments, is shown for context. States are colored by experiment; the same colors are used in panels A–E and G–H and in <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0059865#pone.0059865.s005" target="_blank">Figure S5</a>. Arrows show the directions in which three of the seven state parameters increase. is the probability of reversal, is mean deskewed speed, and is the covariance of deskewed speed and acceleration. A–D: States discovered in four experiments. Lines join states discovered in the same worm. A. 14 wild-type worms, fasted for 12 hours, refed on good food (<i>E coli</i> HB101) for 3 hours, then recorded on good food. B. 12 wild-type worms, grown on good food and recorded on poor food. (Poor food is HB101 treated with aztreonam, which prevents cell division <a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0059865#pone.0059865-BenArous1" target="_blank">[7]</a>.) C. 12 mutant worms lacking cGMP-dependent protein kinase (PKG, encoded in <i>C elegans</i> by <i>egl-4 </i><a href="http://www.plosone.org/article/info:doi/10.1371/journal.pone.0059865#pone.0059865-Fujiwara1" target="_blank">[6]</a>), grown and recorded on good food. D. 12 transgenic worms that express constitutively active PKG in ASI neurons, grown and recorded on good food. E. States from the four previous experiments plotted together. F. Regions of the triangle can be identified as roughly corresponding to roaming, dwelling, and quiescence, as described in the text. G. All behavioral states discovered in 49 experiments on 363 worms. H. States from all experiments on wild-type worms (80 worms total). These experiments differ only in whether the worms were well-fed or starved and refed, and in the quality of food on which they were recorded.