Fig. 4: Results for mapping CCFv3 to sets of hemi-brain BARseq sections, aggregated into 200 μm representations for two separate mice.

a, b Show initial alignment of CCFv3 to BARseq sections (black dots) for a mouse 1 and b mouse 2. c, d Show physically transformed CCFv3 to coordinates of target BARseq sections, with determinant of the Jacobian showing areas of expansion (red) and contraction (blue). e, f Show intersecting sections of deformed CCFv3 (white arrow) overlaid with corresponding BARseq section (black dots) from each hemi-brain. g–l Show intersecting sections of deformed CCFv3 with ontology regions in grayscale and cell type delineations for BARseq sections overlaid in color, with g, h taken at white arrow and i-l more posteriorly (yellow bracket in c, d). m, n Show fraction of correctly aligned cells within each hemi-brain per each of four layer delineations (L2/3, L4, L5, L6) to corresponding CCFv3 region within 500-μm invervals along caudal-rostral axis. Total number of cells within each of four layer delineations given in Supplementary Data 1. o Shows average fraction for 7 hemi-brains of correctly aligned cells with manual alignment (data per hemi-brain provided in Supplementary Table 3) versus fractions achieved with xIV-LDDMM for the two hemi-brains shown.