Clustering mode

Number of clusters (k)

Top N peaks

Resolution bins

Compute all k

Cache every k — enables cycling and Compare tab

Max k

Upper bound on cluster count

Min cluster size

Mark outliers

Size threshold

Clusters smaller than this are outliers

Prominence weight

0 = score · 1 = prominence

Ranking mode

Max k fraction

max_k = ceil(fraction × leaves)

Min prominence

Relative prominence

Ignore k=2

Prefer fewer outliers

Ratio mode

Polytomy mode

No split zero-length

Use branch support

Min support

Support weight

Output directory