Figure 3. | 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 3.

Affiliation details

University of California, Berkeley, United States; Northwestern University, United States; Rehabilitation Institute of Chicago, United States
Figure 3.
Download figureOpen in new tabFigure 3. Model inferences when the true generating model differs from our distance-block-model prior.

Horizontal columns show results with synthetic data generated according to the distance-dependent stochastic block model, the non-distance-dependent stochastic block model, the mixed membership block model, and the latent position cluster model. In all cases histograms represent posterior distribution over the indicated metric. (A) The number of types found by the model; the vertical dashed line indicates the ‘true’ type number (not applicable to the mixed membership model). (B) The area under the receiver operating characteristic (ROC) curve, indicating link prediction accuracy. (C, D, E) Clustering metrics quantifying degree of type agreement with known ground truth.

DOI: http://dx.doi.org/10.7554/eLife.04250.004