unravel.allen_institute.mapmysections.soma_ratio module#

Use mms_soma_ratio or mms_sr from UNRAVEL to compute the proportion of somata voxels in specified atlas regions.

Note

  • We used this to compute the proportion of voxels in the anterior commissure containing somata.

  • This was used to label samples as preferentially having oligodendrocytes.

  • For each seg file, the count of of non-zero voxels in a specified atlas label is divided by the total voxels of that atlas label.

  • If the ratio is > 0.004, the sample is predicted to be oligodendrocyte-enriched.

Prereqs:

  • warp_to_atlas to warp segmentation volumes to atlas space (e.g., CCFv3 2020 30um).

Output:

  • soma_ratio/<input stem>_soma_ratio.csv

  • Columns: label, count, total_label, proportion.

  • label = atlas label (e.g., 890 for anterior commissure)

  • count = count of nonzero voxels in that label

  • total_label = total voxels of that label in the atlas

  • proportion = count / total voxels of that label in the atlas

Next steps:

  • Use mms_concat_with_source to concatenate multiple CSVs into one file.

Usage:#

mms_soma_ratio -a path/atlas_CCFv3_2020_30um.nii.gz [-i ‘<asterisk>_seg_1.nii.gz’] [-l 890] [-o output_dir] [-f] [-v]

unravel.allen_institute.mapmysections.soma_ratio.parse_args()[source]#
unravel.allen_institute.mapmysections.soma_ratio.soma_ratio(seg_path, atlas_img, labels, out_dir, force)[source]#
unravel.allen_institute.mapmysections.soma_ratio.main()[source]#