Coordinate-based meta-analysis (ALE) of fMRI in PTSD stratified by trauma nature and onset: coordinate data, statistical maps and reproducible code (Arm 1, childhood-onset interpersonal trauma, 15 reports; Arm 2, adult-onset trauma, 17 studies)"

<p>Data, statistical maps, figures and reproducible code for an exploratory coordinate-based meta-analysis (Activation Likelihood Estimation, ALE) of task and resting-state fMRI in post-traumatic stress disorder (PTSD), stratified by the nature and onset of the index trauma into two parallel arms.</p><p><strong>Arm 1</strong>: PTSD due to childhood-onset chronic interpersonal trauma (a proxy for ICD-11 complex PTSD); 15 reports, 112 foci. <strong>Arm 2</strong>: PTSD due to adult-onset trauma (discrete or chronic); 17 studies, 186 foci, 276 patients / 254 controls.</p><p><strong>Methods.</strong> NiMARE 0.20.0 ALE; Gaussian kernel with FWHM derived from the patient group size; voxelwise p&lt;0.001 uncorrected (Z&gt;3.0902); cluster-level Monte Carlo FWE p&lt;0.05 by cluster size and by cluster mass; MNI152 space (Talairach converted with Lancaster 2007). Harricharan 2019 and Nicholson 2018 (same Lanius cohort) are pooled in the principal Arm 1 analysis.</p><p><strong>Main results (exploratory).</strong> Arm 1 principal PTSD&gt;HC is null (Z 3.84); a preliminary right inferior-frontal/insular cluster (42, 16, -10; Z 4.93) disappears once the Lanius cohort is pooled and is therefore an artifact of cohort non-independence. Arm 2 PTSD&gt;HC shows a right amygdala/hippocampus cluster (28, -6, -18; Z 4.30; 1152 mm3) surviving size and mass FWE. Both arms are under-powered (k below the ~17-20 experiments required for robust ALE); all findings are hypothesis-generating, not confirmatory.</p><p><strong>Contents.</strong> Peak coordinates (XLSX and CSV, with codebooks), GingerALE/Sleuth foci files, NiMARE dataset JSONs, unthresholded and thresholded statistical maps (NIfTI, MNI152), cluster/jackknife/strata tables, nine figures (PRISMA, glass-brain, orthogonal slices, jackknife, strata), the reproduction code (Python/NiMARE), and the manuscript. No coordinates, statistics or effect sizes were fabricated; missing values are left empty or marked NR. Primary-study PDFs are not redistributed (third-party copyright).</p>"