Atlas-scale single-cell analysis beyond in-memory paradigm with scAtlasPy

Abstract

Single-cell atlases are rapidly outgrowing the memory capacity of standard workstations, challenging the in-memory paradigm underlying mainstream computational ecosystems. Here, scAtlasPy decouples scale of atlas from memory capacity by leveraging the disk-resident computing. It enables full-resolution analysis of a 100-million-cell atlas with only 42.9 GB peak memory, whereas state-of-the-art platforms are limited at 3 million cells with 512 GB memory. scAtlasPy achieves 137,745 cells/s, 10.4x faster than scDataset with 82.6% lower memory usage for random minibatch retrieval. Its extensible architecture offers a flexible platform for diverse atlas-scale analytical tasks, facilitating the discovery of complex cellular heterogeneity and functions in massive cell atlases.

Beyond the Single Pathogen Paradigm in Citrus: A Pathobiome Perspective on Disease Complexes, Microbiome Dysbiosis, and Management
Min Luo, Hongyu Zhang, Ting Zhong +5
Plants (Basel, Switzerland)1 Oct 2026Peer-reviewedCC BYFull textPlant biologyEurope PMC
Emotional Processing in a Nap Paradigm: The Role of Sleep Stages in Memory and Reactivity
Costanza Colombo, Angie Baldassarri, Umberto Granziol +5
Psychophysiology1 Oct 2026Peer-reviewedCC BYFull textNeuroscienceEurope PMC
Astrocytes and memory: emerging evidence and a paradigm shift
Sen Zhang, Qinglu Wang
Cell Communication and Signaling : CCS31 Jul 2026Peer-reviewedCC BY-NC-ND 4.0Full text1 citedNeurosciencePMC
Virtual Silcton and Beyond: A Virtual Environment Paradigm Framework for Spatial Navigation Research
Steven Weisberg, Victor Schinazi
PsyArXiv19 Oct 2026PreprintCC BY 4.0Full textNeuroscienceOSF
Cell identity competes with surface tension to determine stromal organization in tumor spheroids beyond the CAF shell paradigm.
Claire Wilhelm
Proceedings of the National Academy of Sciences of the United States of America12 Oct 2026Peer-reviewedCC BYFull textCell biologyEurope PMC

Cite

BibTeX RIS

BibTeX

@misc{xu2026atlasscale,
  title = {{Atlas-scale single-cell analysis beyond in-memory paradigm with scAtlasPy}},
  author = {H. Xu and Y. Ye and S. Zhang and R. Xie and J. Li and J. Lin and Y. Hu and L. Gao},
  howpublished = {bioRxiv},
  year = {2026},
  month = sep,
  doi = {10.64898/2026.09.03.748766},
  url = {https://www.biorxiv.org/content/10.64898/2026.09.03.748766v1},
  note = {CC BY},
}

RIS

TY  - UNPB
TI  - Atlas-scale single-cell analysis beyond in-memory paradigm with scAtlasPy
AU  - H. Xu
AU  - Y. Ye
AU  - S. Zhang
AU  - R. Xie
AU  - J. Li
AU  - J. Lin
AU  - Y. Hu
AU  - L. Gao
PY  - 2026
DA  - 2026/09/03
PB  - bioRxiv
DO  - 10.64898/2026.09.03.748766
UR  - https://www.biorxiv.org/content/10.64898/2026.09.03.748766v1
AB  - Single-cell atlases are rapidly outgrowing the memory capacity of standard workstations, challenging the in-memory paradigm underlying mainstream computational ecosystems. Here, scAtlasPy decouples scale of atlas from memory capacity by leveraging the disk-resident computing. It enables full-resolution analysis of a 100-million-cell atlas with only 42.9 GB peak memory, whereas state-of-the-art platforms are limited at 3 million cells with 512 GB memory. scAtlasPy achieves 137,745 cells/s, 10.4x faster than scDataset with 82.6% lower memory usage for random minibatch retrieval. Its extensible architecture offers a flexible platform for diverse atlas-scale analytical tasks, facilitating the discovery of complex cellular heterogeneity and functions in massive cell atlases.
ER  -

Abstract

Related

Cite