Preserve full GDC biospecimen hierarchy (samples/portions/analytes/aliquots); add TcgaHfPatient pydantic reference impl

README.md

@@ -25,7 +25,7 @@ configs:
 One HuggingFace subset per TCGA project; one row per patient.
 
 - **Projects included:** `TCGA-CHOL`, `TCGA-DLBC`
-- **Generated:** 2026-05-04 03:00:27 UTC
+- **Generated:** 2026-05-04 03:21:55 UTC
 - **Source:** NCI GDC `/cases` endpoint, open-access tier only.
 - **GDC data release:** Data Release 45.0 - December 04, 2025
 
@@ -77,6 +77,20 @@ Normal`). Each variant carries the aliquot UUIDs of both members of the pair.
   unambiguous.
 - `experimental_strategy`: the underlying assay (e.g. *WXS*, *RNA-Seq*).
 
+**Reference Python implementation:** the `tcga2hf` package on GitHub ships a
+fully-typed pydantic `TcgaHfPatient` model that mirrors this schema and adds
+convenience joins (tumor/normal pairs, mutations-by-gene, expression-by-gene,
+longitudinal timeline). Useful both as a loader and as documentation for what
+the data means.
+
+```python
+import pyarrow.parquet as pq
+from tcga2hf.models import TcgaHfPatient
+
+t = pq.read_table("TCGA-CHOL/train.parquet")
+patients = [TcgaHfPatient.model_validate(r) for r in t.to_pylist()]
+```
+
 **Authoritative GDC references:**
 - [Data dictionary][gdc-dict] (every entity + field definition)
 - [Biospecimen Encyclopedia](https://docs.gdc.cancer.gov/Encyclopedia/pages/Biospecimen/)

TCGA-CHOL/train.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:fafc0ffd0051d84022e214edfa7d95ebed2c5b44dfd1943ea23dee229d5fee9a
-size 51274704
+oid sha256:b261c92f2a1d02c1412cbb1b9fdc381c8a9c9e0e4967ee93f1547f1b8702da1f
+size 51303188

TCGA-DLBC/train.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:019961cbfc1a65d1009f423ae69a61888bc5bf8a73e64e0d1910ff01ee1166b1
-size 57738136
+oid sha256:c08c0c1eaae44aa1fc0de45a6c6002b1f99a0ebdfe321081176eea16afc92002
+size 57772197