Fig. 2
From: A mixture model for signature discovery from sparse mutation data
A plate diagram for Mix. Notation is the same as in Fig. 1. The key difference from MMM as illustrated in the plate diagram is that the (latent) selected signature \(z_{t}^{n}\) depends on a cluster cn chosen for each of the samples (with prior probabilities w), as well as (latent) cluster exposures πl shared across samples within that cluster