Simplify figures and move explanations into the dataset card

README.md

@@ -13,9 +13,9 @@ configs:
 ---
 # SpatialDataAgent-DB Public Catalog
 
-![SpatialDataAgent-DB overview with real H&E, GFAP, DAPI and CosMx previews](assets/dataset-card/v1/01-overview.png)
+![H&E, spatial GFAP, DAPI and CosMx atlas previews](assets/dataset-card/v2/01-overview.png)
 
-*The selected source snapshot contains 12,163 samples and 635.24 million spatial observations. H&E and GFAP are paired views of GSM9721856; the expression panel shows all 3,416 recorded spots. DAPI and CosMx previews illustrate additional image modalities.*
+*Left to right: H&E and GFAP from the same Visium section (GSM9721856), Xenium DAPI (GSM9301054), and a CosMx composite (GSM9372712). The GFAP panel displays all 3,416 recorded spots using the existing log1p preview values.*
 
 Search and inspect the metadata for 12,163 samples, 1,265 published study folders and 635,240,170 spatial observations in the fixed `sda-unified-20261010-v3` snapshot. It includes 8,211 samples with released tissue images.
 
@@ -29,12 +29,10 @@ Source index SHA256: `909c4ec50e71c636eadc7bf01a9479fef1b4fd480ac348dfea6b785e56
 
 ## Collection composition
 
-![All-sample platform counts, image availability, source collections and frozen splits](assets/dataset-card/v1/02-composition.png)
+![Sample counts by platform, image availability, source collection and split](assets/dataset-card/v2/02-composition.png)
 
-*Platform counts, image availability, source collections and retained study-level splits use all 12,163 samples in the frozen source snapshot.*
+*Sample counts by platform, image availability, source collection and retained split (n = 12,163; source snapshot: sda-unified-20261010-v3).*
 
 ## Access and reuse
 
-![Public catalog, manual approval, full-sample download and analysis workflow](assets/dataset-card/v1/03-access-and-reuse.png)
-
-*Explore public metadata, request file access, retrieve selected samples and analyse them. Streaming training shards and the Python SDK are planned follow-up outputs.*
+Browse the public catalog, request access to the [full sample data](https://huggingface.co/datasets/Biogod/SpatialDataAgent-DB), then retrieve H5AD, available TIFF images and JSON metadata. AnnData/Scanpy can read downloaded expression files. Streaming training shards and the Python SDK are planned follow-up outputs.

assets/dataset-card/v2/verification.json

@@ -0,0 +1,48 @@
+{
+  "source_release": "sda-unified-20261010-v3",
+  "checks": [
+    "Sample denominator: 12,163",
+    "Spatial observations: 635,240,170",
+    "Unchanged source catalog binding",
+    "GFAP preview matches catalog expression SHA256",
+    "All 3,416 GFAP spots retained; values and coordinates unchanged",
+    "01-overview: labels within canvas []",
+    "Platform counts match frozen catalog",
+    "Image count: 8,211",
+    "Source counts match frozen catalog",
+    "Split counts match frozen catalog",
+    "02-composition: labels within canvas []"
+  ],
+  "products": [
+    {
+      "name": "01-overview.png",
+      "bytes": 1453502,
+      "sha256": "5a1cbd439ee1985a0388b7acb83e5c58fb64859b70f51dc3426108e93676987d"
+    },
+    {
+      "name": "01-overview.svg",
+      "bytes": 1855769,
+      "sha256": "e472dddc15da9b1336f39ad70e174a0696aaa718c498c32264cd449f4ad83cc4"
+    },
+    {
+      "name": "02-composition.png",
+      "bytes": 119478,
+      "sha256": "51f48eb451c2cd314ed27a0c0508991f96d4bad3b0a8219ba2ec2e6c5facfb0b"
+    },
+    {
+      "name": "02-composition.svg",
+      "bytes": 16611,
+      "sha256": "57b0f6efda282e77e823285f0c8bc8ac81f9e937a0b7b31c5277a38345322260"
+    }
+  ],
+  "source_figures": "../figures-v1",
+  "image_samples": [
+    "GSM9721856",
+    "GSM9301054",
+    "GSM9372712"
+  ],
+  "tissue_pixels_edited": false,
+  "generated_tissue_images": false,
+  "design_changes": "No overall titles, statistics cards, captions, footnotes or workflow image; explanations live in Markdown.",
+  "matplotlib_version": "3.11.2"
+}