Datasets:
cell_id stringlengths 10 10 | vertex_x float32 3.79k 5.32k ⌀ | vertex_y float32 3.08k 4.6k ⌀ | label_id int64 101k 719k ⌀ |
|---|---|---|---|
capakghg-1 | 5,297.218262 | 4,213.933105 | 101,288 |
capakghg-1 | 5,297.005859 | 4,214.145508 | 101,288 |
capakghg-1 | 5,294.668457 | 4,214.145508 | 101,288 |
capakghg-1 | 5,294.243652 | 4,214.570313 | 101,288 |
capakghg-1 | 5,294.243652 | 4,215.42041 | 101,288 |
capakghg-1 | 5,293.605957 | 4,216.058105 | 101,288 |
capakghg-1 | 5,293.605957 | 4,216.908203 | 101,288 |
capakghg-1 | 5,294.030762 | 4,217.333008 | 101,288 |
capakghg-1 | 5,294.668457 | 4,218.183105 | 101,288 |
capakghg-1 | 5,295.306152 | 4,218.183105 | 101,288 |
capakghg-1 | 5,295.943359 | 4,218.60791 | 101,288 |
capakghg-1 | 5,297.005859 | 4,218.60791 | 101,288 |
capakghg-1 | 5,297.643555 | 4,219.245605 | 101,288 |
capakghg-1 | 5,297.855957 | 4,219.032715 | 101,288 |
capakghg-1 | 5,298.706055 | 4,219.032715 | 101,288 |
capakghg-1 | 5,299.980957 | 4,217.757813 | 101,288 |
capakghg-1 | 5,300.193359 | 4,217.333008 | 101,288 |
capakghg-1 | 5,300.193359 | 4,216.695313 | 101,288 |
capakghg-1 | 5,300.405762 | 4,216.48291 | 101,288 |
capakghg-1 | 5,300.405762 | 4,215.208008 | 101,288 |
capakghg-1 | 5,299.768066 | 4,214.570313 | 101,288 |
capakghg-1 | 5,298.918457 | 4,214.145508 | 101,288 |
capakghg-1 | 5,297.855957 | 4,214.145508 | 101,288 |
capakghg-1 | 5,297.643555 | 4,213.933105 | 101,288 |
capakghg-1 | 5,297.218262 | 4,213.933105 | 101,288 |
capanikj-1 | 5,279.793945 | 4,253.244141 | 101,289 |
capanikj-1 | 5,279.581543 | 4,253.456543 | 101,289 |
capanikj-1 | 5,278.943848 | 4,253.456543 | 101,289 |
capanikj-1 | 5,278.731445 | 4,253.668945 | 101,289 |
capanikj-1 | 5,278.519043 | 4,253.668945 | 101,289 |
capanikj-1 | 5,277.668945 | 4,254.519043 | 101,289 |
capanikj-1 | 5,277.668945 | 4,256.006348 | 101,289 |
capanikj-1 | 5,277.456543 | 4,256.219238 | 101,289 |
capanikj-1 | 5,277.456543 | 4,257.918945 | 101,289 |
capanikj-1 | 5,278.306641 | 4,258.769043 | 101,289 |
capanikj-1 | 5,278.519043 | 4,258.769043 | 101,289 |
capanikj-1 | 5,278.731445 | 4,258.981445 | 101,289 |
capanikj-1 | 5,280.006348 | 4,258.981445 | 101,289 |
capanikj-1 | 5,280.21875 | 4,258.769043 | 101,289 |
capanikj-1 | 5,280.431641 | 4,258.769043 | 101,289 |
capanikj-1 | 5,282.556641 | 4,256.856445 | 101,289 |
capanikj-1 | 5,283.831543 | 4,255.581543 | 101,289 |
capanikj-1 | 5,283.831543 | 4,255.369141 | 101,289 |
capanikj-1 | 5,284.043945 | 4,255.156738 | 101,289 |
capanikj-1 | 5,284.043945 | 4,254.731445 | 101,289 |
capanikj-1 | 5,283.618652 | 4,254.306641 | 101,289 |
capanikj-1 | 5,282.981445 | 4,253.456543 | 101,289 |
capanikj-1 | 5,282.556641 | 4,253.456543 | 101,289 |
capanikj-1 | 5,282.34375 | 4,253.244141 | 101,289 |
capanikj-1 | 5,279.793945 | 4,253.244141 | 101,289 |
capannac-1 | 5,287.443848 | 4,424.937988 | 101,290 |
capannac-1 | 5,286.806152 | 4,425.150391 | 101,290 |
capannac-1 | 5,285.956055 | 4,426.000488 | 101,290 |
capannac-1 | 5,285.956055 | 4,426.425293 | 101,290 |
capannac-1 | 5,284.894043 | 4,427.487793 | 101,290 |
capannac-1 | 5,285.106445 | 4,428.125 | 101,290 |
capannac-1 | 5,284.894043 | 4,428.550293 | 101,290 |
capannac-1 | 5,286.806152 | 4,430.037598 | 101,290 |
capannac-1 | 5,287.443848 | 4,430.887695 | 101,290 |
capannac-1 | 5,289.143555 | 4,432.375 | 101,290 |
capannac-1 | 5,289.78125 | 4,432.587402 | 101,290 |
capannac-1 | 5,290.84375 | 4,433.4375 | 101,290 |
capannac-1 | 5,291.268555 | 4,433.4375 | 101,290 |
capannac-1 | 5,292.118652 | 4,433.012695 | 101,290 |
capannac-1 | 5,292.543457 | 4,432.587402 | 101,290 |
capannac-1 | 5,292.755859 | 4,430.675293 | 101,290 |
capannac-1 | 5,292.755859 | 4,429.825195 | 101,290 |
capannac-1 | 5,292.543457 | 4,428.975098 | 101,290 |
capannac-1 | 5,292.118652 | 4,428.125 | 101,290 |
capannac-1 | 5,292.331055 | 4,427.275391 | 101,290 |
capannac-1 | 5,289.993652 | 4,425.362793 | 101,290 |
capannac-1 | 5,289.568848 | 4,425.150391 | 101,290 |
capannac-1 | 5,288.506348 | 4,425.150391 | 101,290 |
capannac-1 | 5,288.293457 | 4,424.937988 | 101,290 |
capannac-1 | 5,287.443848 | 4,424.937988 | 101,290 |
capapapc-1 | 5,279.156738 | 4,301.267578 | 101,291 |
capapapc-1 | 5,278.519043 | 4,301.47998 | 101,291 |
capapapc-1 | 5,277.456543 | 4,302.967285 | 101,291 |
capapapc-1 | 5,277.668945 | 4,303.39209 | 101,291 |
capapapc-1 | 5,277.456543 | 4,303.60498 | 101,291 |
capapapc-1 | 5,278.094238 | 4,304.242188 | 101,291 |
capapapc-1 | 5,279.581543 | 4,304.242188 | 101,291 |
capapapc-1 | 5,279.793945 | 4,304.45459 | 101,291 |
capapapc-1 | 5,279.581543 | 4,305.092285 | 101,291 |
capapapc-1 | 5,280.856445 | 4,305.942383 | 101,291 |
capapapc-1 | 5,282.556641 | 4,306.154785 | 101,291 |
capapapc-1 | 5,284.46875 | 4,307.004883 | 101,291 |
capapapc-1 | 5,285.956055 | 4,307.004883 | 101,291 |
capapapc-1 | 5,286.168945 | 4,307.217285 | 101,291 |
capapapc-1 | 5,288.081055 | 4,307.004883 | 101,291 |
capapapc-1 | 5,289.143555 | 4,306.367188 | 101,291 |
capapapc-1 | 5,289.568848 | 4,305.304688 | 101,291 |
capapapc-1 | 5,288.506348 | 4,304.029785 | 101,291 |
capapapc-1 | 5,286.381348 | 4,302.967285 | 101,291 |
capapapc-1 | 5,285.53125 | 4,302.967285 | 101,291 |
capapapc-1 | 5,284.256348 | 4,302.754883 | 101,291 |
capapapc-1 | 5,282.769043 | 4,302.54248 | 101,291 |
capapapc-1 | 5,281.28125 | 4,301.692383 | 101,291 |
capapapc-1 | 5,279.369141 | 4,301.267578 | 101,291 |
capapapc-1 | 5,279.156738 | 4,301.267578 | 101,291 |
Cervical Xenium WTA — B-follicle / TLS workshop crop
A small, self-contained spatial crop of the Atera 10x Xenium WTA "Cervical Cancer FFPE" slide, centred on the best B-cell-follicle / tertiary-lymphoid- structure (TLS) aggregate on the section. Built for hands-on workshop use: it carries every native Xenium data layer (cells, molecules, segmentation, morphology image) at full resolution but is ~1/35th of the slide area, so it downloads in seconds.
Full dataset: this is a crop of the public 10x Genomics Atera WTA FFPE Human Cervical Cancer slide (717,576 cells, 18,028 genes, CC BY 4.0). Download the whole slide there.
Focal aggregate
The densest B-cell follicle on the slide
(results/cervical_aggregate_scores.csv), centroid (4380.72, 3664.42) µm:
n_B = 96, n_plasma = 23, CXCL13⁺ = 171, n_T = 228, n_DC = 24. This is the most
mature-TLS-like aggregate by follicle-core architecture (B follicle + plasma
cells + germinal-centre CXCL13 signal wrapped in a T zone). See the case study
reports/case_study_atera_cervical_triads_tls.md for the analysis context.
Bounding box (slide micron coordinates)
The 1500 µm square box places the follicle centroid at the golden-section point (0.382 of the side from the min corner on each axis), so the larger context margin falls toward +x / +y.
| axis | min | max | extent |
|---|---|---|---|
| x | 3807.8 µm | 5307.8 µm | 1500 µm |
| y | 3091.5 µm | 4591.5 µm | 1500 µm |
Same coordinate frame as the parent cervical_cancer.h5ad
(obs.x_centroid / obs.y_centroid, units = µm, origin = slide top-left).
All files below keep absolute slide microns — nothing is re-origined — so
an analysis written for the full slide runs unchanged on this crop.
Files
| file | contents |
|---|---|
cervical_tls_crop.h5ad |
26,047 cells × 18,028 genes, CSR X, full parent obs + region / is_triad / is_anchor / lineage / cell_type_major merged from results/triad_membership_per_cell.parquet. obsm['spatial'] and obs.x_centroid / obs.y_centroid are absolute slide microns. |
cell_boundaries.parquet |
650,742 cell-segmentation polygon vertices (cell_id, vertex_x, vertex_y, label_id) for the kept cells, in absolute slide microns. |
nucleus_boundaries.parquet |
649,439 nucleus polygon vertices, same schema. |
transcripts.parquet |
48,650,149 per-molecule detections whose x_location / y_location fall in the box. Columns: feature_name, cell_id, x_location, y_location, z_location, qv, overlaps_nucleus, nucleus_distance, is_gene (the four Xenium decoding-internal columns — transcript_id, fov_name, codeword_index, codeword_category — are dropped to slim the download; is_gene + feature_name already separate genes from negative-control codewords). Includes UNASSIGNED molecules (not in any cell) and control-probe detections, as Xenium emits them. |
morphology_focus_crop.ome.tif |
4-channel pyramidal OME-TIFF (DAPI, ATP1A1/CD45/E-Cadherin, 18S, αSMA/Vimentin), uint16, tiled 1024², JPEG2000-lossless (Xenium's native codec), 8 pyramid levels. L0 = 7059 × 7059 px @ 0.21249222 µm/px. Crop L0 pixel (0,0) corresponds to slide µm (3807.86, 3091.55). |
verification_overlay.png |
QC thumbnail: DAPI + cell centroids coloured by immune lineage, white rings on triad cells, red + at the follicle centroid. |
Overlaying cells / molecules on the morphology crop
obs.x_centroid, cell_boundaries.vertex_x, and transcripts.x_location are
all absolute slide microns. The crop image L0 pixel (0, 0) is at slide µm
(3807.86, 3091.55). To convert to crop pixels:
PIXEL_UM = 0.21249222
OX, OY = 3807.860582, 3091.549309 # crop origin in slide µm
crop_px_x = (x_um - OX) / PIXEL_UM
crop_px_y = (y_um - OY) / PIXEL_UM
Cell counts
| label | cells |
|---|---|
| total | 26,047 (3.63 % of slide's 717,576) |
| is_triad | 237 |
| is_anchor | 25 |
Provenance
Built by notebooks/atera_xenium_wta/scripts/build_tls_workshop_crop.py from:
data/atera/WTA_Preview_FFPE_Cervical_Cancer_outs/cervical_cancer.h5ad.../cell_boundaries.parquet,.../nucleus_boundaries.parquet.../morphology_focus/ch0000_dapi.ome.tif(+ sibling channels)transcripts.parquet(extracted fromWTA_Preview_FFPE_Cervical_Cancer_outs.zip)notebooks/atera_xenium_wta/results/triad_membership_per_cell.parquet- focal aggregate from
notebooks/atera_xenium_wta/results/cervical_aggregate_scores.csv
- Downloads last month
- 57