Custom Weights

Clustering now supports custom weights to mutations as given in a column from the input .maf. Users may now predefine how each mutation contributes to the closeness centrality, in contrast to the recurrence calculation which equivalently assumes a weight of 1 for each sample with a mutation. When using vertex-type == weight, the specified weight-header column from the .maf will be read in as the value specified by the user's annotation.

Cluster significance

A new module to calculate cluster significance (hotspot3d sigclus) is added. Determine cluster p-values from permutation tests. This helps improve cluster interpretation.

Recursive clustering at hotspots

Added recursive clustering of the agglomerative hierarchical clustering method.