Pancreatic Ductal Adenocarcinoma (PDAC) is heterogeneous with low tumor purity, prominent microenvironment and complex architecture, which preclude identification of shared tumor intrinsic biology within and across patients. We overcame these challenges by applying spatial omics approaches – providing necessary resolution and context – to primary untreated PDAC tumors from 39 patients capturing 340K low-bulk and 530K single-cell spatial transcriptomes. We observe a spectrum of classical-to-basal tumor subtypes present within all patients. We identify a distinct population of proliferative tumor cells enriched for the DNA synthesis-condensation-mitosis transcriptional network suggesting this subset disproportionately contribute to tumor growth. Furthermore, we characterize the microenvironment surrounding each tumor subtype and observe that basal identity is contextually defined by collagen, hypoxic adaptation and fibroblast neighbors expressing activated stroma and interferon stimulated gene therapy resistance signatures. Lastly, leveraging orthogonal patient, cell and mouse data, we determine which aspects of PDAC biology are recapitulated in bulk datasets and reductionist models.
from sddb import Catalog
cohort = Catalog().collection("geo-gse282302")
cohort.members # the 108 datasets as a table
print(cohort.licenses()) # check the terms before you download
sdatas = cohort.to_dict() # {uid: SpatialData}, cached locally
# or keep the .zarr files in a folder of your choice:
# sdatas = cohort.to_dict(dest="data/")Needs pip install lamindb and network access to the Lamin database, which is limited during the alpha. If it fails, use the Client tab.
import lamindb as ln
db = ln.DB("scverse/spatialdata-db")
coll = db.Collection.get(key="geo-gse282302")
coll.ordered_artifacts.to_dataframe() # the 108 members, ordered
# download every member .zarr to the local cache:
paths = [a.cache() for a in coll.ordered_artifacts.all()]# sddb client: download every member .zarr into data/
sddb-cli download --collection geo-gse282302 --dest data/Licenses: 108 NCBI GEO policy
@article{Lyubetskaya_2026,
title = {In situ multi-modal characterization of pancreatic cancer reveals tumor cell identity as a defining factor of the surrounding microenvironment},
volume = {45},
ISSN = {2211-1247},
url = {http://dx.doi.org/10.1016/j.celrep.2025.116827},
DOI = {10.1016/j.celrep.2025.116827},
number = {1},
journal = {Cell Reports},
publisher = {Elsevier BV},
author = {Lyubetskaya, Anna and Rabe, Brian and Kavran, Andrew and Bai, Yulong and Fisher, Andrew and Font-Tello, Alba and Lewin, Anne and Pliner, Hannah and Fan, Yunfan and Giampapa, Lauren and Cheng, Yelena and Dai, Chao and Hu, Ruifeng and Lila, Tom and Alloy, Alexandre P. and Mason, Mike and Brett, Constance and Brett, Todd and Seifuddin, Fayaz and Vasquez Grinnell, Steven and Stahlschmidt, Soren R. and Zhao, Yilin and Golhar, Ryan and Neuhaus, Isaac and Carrera, Daniel and Rios, Carlos and Kar, Pradeep and Shukla, Abhishek and Bashford-Rogers, Rachael and Meyer, Matthew J. and Abu Shah, Enas and Heij, Lara and Sivakumar, Shivan and Trillo-Tinoco, Jimena and Chen, Benjamin J. and Mavrakis, Konstantinos J. and Drokhlyansky, Eugene},
year = {2026},
month = Jan,
pages = {116827}
}