Figure 7. | Automatic discovery of cell types and microcircuitry from neural connectomics

Open accessCopyright infoDownload PDF

Automatic discovery of cell types and microcircuitry from neural connectomics

Figure 7.

Affiliation details

University of California, Berkeley, United States; Northwestern University, United States; Rehabilitation Institute of Chicago, United States
Figure 7.
Download figureOpen in new tabFigure 7. Discovering connectivity and type in C. elegans.

(A) Posterior distribution on cell connectivity as a function of discovered type, similar to Figure 6. In (B) we plot neuroanatomist-derived types along with their labels. Our model shows a high probability of motor neurons, sensory neurons, and various interneuron classes being of the same type. Soma positions along the body axis are plotted in (C) where we see that we cluster spatially distributed motor neurons together, whereas head sensory neurons are more likely to be grouped together as well. (D) The receiver operating characteristic (ROC) curves for held-out link probability for both the electrical synapses (gap junctions) and chemical synapses in C. elegans. (E) The posterior distribution of the area under the ROC curve (AUC) for the curves in (D). (F) Measurements of the agreement of our identified cell types compared to neuroanatomists. The high completeness but low homogeneity (and corresponding low adjusted Rand index, ARI) reflects our model's tendency to group multiple types into a single type.

DOI: http://dx.doi.org/10.7554/eLife.04250.008