Node-wise Tract Explorer

The Node-wise Tract Explorer is a browser-based viewer for node-wise statistical results distributed with this site. It reads a results file in the browser, without uploading it, and presents each analysis as a table row that expands to show the node-wise t-value profile, the clusters identified, and, where both hemispheres are present, a left–right comparison. Whole-tract and quartile results are tabular and do not require it.

Open the Explorer · sample file (synthetic data in the expected format)

Input format

The input is a long-format CSV with one row per node. Six columns are required: outcome, tract, metric, node, t and p. Optional columns are hemisphere (L or R), N, covariates, extent_threshold, cluster_p and passed. Any additional column is treated as a grouping variable and becomes a filter. The column names produced by 09b_nodewise_permutation.R (Node, t_value, p_value) are accepted without renaming. 09c_stack_for_explorer.py (Step 9) produces this file from the permutation outputs.

Derived quantities

From the node rows the viewer derives the number of significant nodes, clusters (contiguous runs with p < .05), the largest cluster, which clusters exceed the extent threshold, and, when a hemisphere column is present, the counts of significant nodes per hemisphere and their overlap.

Display

Selecting a row shows the t-value profile with a labelled axis and reference lines at ±2, significant nodes highlighted, retained clusters shaded, and a cluster table. For lateralized data the left and right profiles are drawn on a common axis so that both hemispheres' full t-value profiles are visible. The hemisphere panel is descriptive: it reports the number of significant nodes per side and their overlap and does not constitute a test of lateralization.

Page labels may be supplied in the URL, for example ?title=…&method=Freedman–Lane&node0=VTA&node1=hippocampus&n_perms=5000 for a ventral tegmental area (VTA) → hippocampus analysis, and ?data=URL loads a hosted file directly.