Methods for validating the legitimacy of clusters using classification. SVMs are used as the basis for the classification. Merging is done based on the connectivity from an SNN graph.
ValidateClusters(object, pc.use = NULL, top.genes = 30,
min.connectivity = 0.01, acc.cutoff = 0.9, verbose = TRUE)Seurat object
Which PCs to use to define genes in model construction
Use the top X genes for each PC in model construction
Threshold of connectedness for comparison of two clusters
Accuracy cutoff for classifier
Controls whether to display progress and merging results
Returns a Seurat object, object@ident has been updated with new cluster info