Gene Co-presence Heatmap and Clustering

Explore which flagellar genes tend to be co-present in the same genomes, and how those genes cluster when genomes are compared by shared presence. The heatmap uses weighted Jaccard similarity; the dendrogram uses complete linkage on 1 − similarity.

Controls and options

Only selected genes are used in the heatmap and clustering. Search the list or use Select All / Deselect All.
—

No genes selected. Search or use Select All for the co-presence heatmap.

Which taxonomy column groups genomes (e.g. species vs. order). Finer ranks make more, smaller groups; coarser ranks pool lineages for correcting sampling bias. Pairs with Normalization factor, which adjusts weights within each group.
none — every genome counts the same.inverse — big taxon groups contribute less per genome (weight ∝ 1 ÷ group size).inverse_sqrt — for correcting sampling bias with a weaker factor (weight ∝ 1 ÷ √(group size)).

Similarity scale (colors)

Similarity runs from your low color through white at 0.5 to your high color. Tweak for contrast or to match other figures.
Low similarity
High similarity

Clustered Gene Co-presence Heatmap

Gene Co-presence Network

Force-directed co-presence network: genes are nodes, and edges link pairs whose Jaccard similarity meets the threshold (optionally capped to the strongest edges per node). Positions come from a physics simulation—stronger associations pull genes together while repulsion and collision keep the layout readable. Drag nodes to rearrange; scroll to zoom.

0.50

0 genes · 0 edges shown

Select at least one gene to build the co-presence network.