Deep generative AI models analyzing circulating orphan non-coding RNAs enable detection of early-stage lung cancer
Liquid biopsies have the potential to revolutionize cancer care through non-invasive early detection of tumors. Developing a robust liquid biopsy test requires collecting high-dimensional data from a large number of blood samples across heterogeneous groups of patients. We propose that the generative capability of variational auto-encoders enables learning a robust and generalizable signature of blood-based biomarkers. In this study, we analyze orphan non-coding RNAs (oncRNAs) from serum samples of 1050 individuals diagnosed with non-small cell lung cancer (NSCLC) at various stages, as well as sex-, age-, and BMI-matched controls. We demonstrate that our multi-task generative AI model, Orion, surpasses commonly used methods in both overall performance and generalizability to held-out datasets. Orion achieves an overall sensitivity of 94% (95% CI: 87%-98%) at 87% (95% CI: 81%-93%) specificity for cancer detection across all stages, outperforming the sensitivity of other methods on held-out validation datasets by more than ~ 30%.
Authors
- Lisa Fish (ORCID: https://orcid.org/0000-0001-9184-0512)
- Fereydoun Hormozdiari (ORCID: https://orcid.org/0000-0003-2703-9274)
- James Zou (ORCID: https://orcid.org/0000-0001-8880-4764)
- Seda Kilinc
- Alice Huang (ORCID: https://orcid.org/0000-0002-4795-3402)
- Jeremy Ku
- Babak Alipanahi (ORCID: https://orcid.org/0000-0001-8216-7178)
- Patrick Arensdorf
- Babak Behsaz (ORCID: https://orcid.org/0000-0003-3313-0322)
- Michael Multhaup
- Anna Hartwig
- Nae-Chyun Chen (ORCID: https://orcid.org/0000-0002-4140-4568)
- Mehran Karimzadeh (ORCID: https://orcid.org/0000-0002-7324-6074)
- Olivier Elemento (ORCID: https://orcid.org/0000-0002-8061-9617)
- Kathleen Wang
- Amir Momen-Roknabadi (ORCID: https://orcid.org/0000-0001-6233-1401)
- Taylor B. Cavazos (ORCID: https://orcid.org/0000-0003-3537-2608)
- Jieyang Wang
- Yuqi Fang (ORCID: https://orcid.org/0000-0001-6750-8536)
- Ti Lam
- Hani Goodarzi (ORCID: https://orcid.org/0000-0002-9648-8949)
- Jennifer Yen (ORCID: https://orcid.org/0000-0002-8580-220X)
- Philip Murzynowski
- Xuan Zhao (ORCID: https://orcid.org/0009-0002-2020-752X)
- Helen Li (ORCID: https://orcid.org/0000-0002-1145-6527)
- Dang Le Tri Nguyen (ORCID: https://orcid.org/0000-0002-7933-4294)
- Rose Hanna
- Diana Corti
- Kimberly H. Chau
Institutions
- Serviço Nacional de Aprendizagem Industrial (BR)
- University of California, San Francisco (US)
- Cornell University (US)
- Science Exchange (United States) (US)
- Arc of the United States (US)
- Weill Cornell Medicine (US)
- University of California, Davis (US)
- Stanford University (US)
Publication Details
- Journal
- Nature Communications
- Published
- 2024-11-21
- DOI
- https://doi.org/10.1038/s41467-024-53851-9
- Citations
- 32
- Primary Topic
- Cancer Genomics and Diagnostics
- Type
- article
- Field-Weighted Citation Impact
- 9.12