Fig. 1From: Latent dirichlet allocation for double clustering (LDA-DC): discovering patients phenotypes and cell populations within a single Bayesian frameworkThe proposed pipeline to perform the double clustering: from the observations (where one patient is represented by a matrix) to the conditional probability distributions of clusters given patients and cellular types given clusters. N is the number of fluorescent markers, K is the number of clusters, W is the number of wordsBack to article page