Fig. 1: Transcriptomic specificity of neuroanatomical effects.

a Schematic outlining the main imaging-transcriptomic enrichment analyses and statistical tests. b (left) Surface projections of T-statistics (z-scored for plotting purposes) for CNV effects on regional morphometric similarity (MS). Despite some overlap across CNVs, each CNV induces a distinct profile of MS change. For full chromosome CNVs, neighboring point range plots show the median (point) and standard error (range) rank of each chromosomal gene set—based on gene rankings from the PLS analysis (see (a), N genes per chromosome provided in Supplementary Dataset 3). The chromosomal gene set for each CNV possessed a more extreme median rank than all other chromosomal gene sets, and the polarity of this effect was opposite for chromosomal duplications (CNV gene set high ranked) versus deletion (CNV gene set low ranked). For subchromosomal CNVs (depicted as red in the respective chromosome ideograms), density plots show median (solid line) and standard error (dashed line) ranks for the relevant CNV gene set. Observed ranks are shown relative to two null distributions: P_RAND-Trans (black), and P_RAND-Cis (gray). P_RAND was calculated using 10,000 gene rank permutations (black). P_RAND-Cis was calculated similarly to P_RAND-Trans but only sampling gene ranks from the respective chromosome of the CNV. All permuted P values were not further corrected for multiple comparisons, and were determined based on one-sided tests of gene set enrichment (median rank; see “Methods”).