Mark Release v0.1.0 (partial)
Browse files
README.md
CHANGED
|
@@ -26,6 +26,13 @@ configs:
|
|
| 26 |
|
| 27 |
# CellImageNet
|
| 28 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 29 |
**CellImageNet** is a large-scale single-cell image database of **paired DAPI
|
| 30 |
nuclear images with cell-type annotations**, built from publicly available
|
| 31 |
10x Genomics Xenium data. It contains **~10 million cells** from **42 Xenium
|
|
@@ -64,6 +71,26 @@ ex = next(iter(ds))
|
|
| 64 |
ex["2p5x.png"], ex["10x.png"], ex["json"]["cell_type"]
|
| 65 |
```
|
| 66 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 67 |
## Sample schema (WebDataset)
|
| 68 |
|
| 69 |
Each sample (one cell) is keyed by `cell_id` with three members:
|
|
|
|
| 26 |
|
| 27 |
# CellImageNet
|
| 28 |
|
| 29 |
+
> ⚠️ **Release v0.1.0 (partial).** This is an initial partial release. The full
|
| 30 |
+
> corpus is **42 Xenium samples (28 human + 14 mouse)**; additional tissues are
|
| 31 |
+
> being uploaded as they finish processing. The figures below describe the
|
| 32 |
+
> **complete** CellImageNet — the set currently available on the Hub is a subset
|
| 33 |
+
> and is growing toward the full 42. See **[Release status](#release-status)**
|
| 34 |
+
> for exactly what is uploaded right now.
|
| 35 |
+
|
| 36 |
**CellImageNet** is a large-scale single-cell image database of **paired DAPI
|
| 37 |
nuclear images with cell-type annotations**, built from publicly available
|
| 38 |
10x Genomics Xenium data. It contains **~10 million cells** from **42 Xenium
|
|
|
|
| 71 |
ex["2p5x.png"], ex["10x.png"], ex["json"]["cell_type"]
|
| 72 |
```
|
| 73 |
|
| 74 |
+
## Release status
|
| 75 |
+
|
| 76 |
+
**v0.1.0 (partial)** — the following samples are currently uploaded; the rest are
|
| 77 |
+
being added as they finish processing (target: 28 human + 14 mouse = 42).
|
| 78 |
+
|
| 79 |
+
Currently available (9 human samples):
|
| 80 |
+
|
| 81 |
+
- Xenium_Preview_Human_Lung_Cancer
|
| 82 |
+
- Xenium_Preview_Human_Non_diseased_Lung
|
| 83 |
+
- Xenium_Prime_Ovarian_Cancer_FFPE
|
| 84 |
+
- Xenium_V1_FFPE_Human_Brain_Alzheimers
|
| 85 |
+
- Xenium_V1_FFPE_Human_Brain_Glioblastoma
|
| 86 |
+
- Xenium_V1_FFPE_Human_Brain_Healthy
|
| 87 |
+
- Xenium_V1_hColon_Cancer_Add_on
|
| 88 |
+
- Xenium_V1_hColon_Cancer_Base
|
| 89 |
+
- Xenium_V1_hColon_Non_diseased_Add_on
|
| 90 |
+
|
| 91 |
+
The authoritative, always-current list of source samples (with 10x URLs and
|
| 92 |
+
per-sample cell counts) is in [`attribution_manifest.csv`](attribution_manifest.csv).
|
| 93 |
+
|
| 94 |
## Sample schema (WebDataset)
|
| 95 |
|
| 96 |
Each sample (one cell) is keyed by `cell_id` with three members:
|