Reducing demographic bias in biomedical machine learning for cancer detection using cfDNA methylation

Shuo Li¹, Weihua Zeng², Wenyuan Li², Chun-Chi Liu³, Yonggang Zhou², Xiaohui Ni³, Mary L Stackpole³, Angela H Yeh², Andrew Melehy⁴, David S Lu^{5

6}, Steven S Raman^{4

5

6}, William Hsu^{5

6}, Lopa Mishra^{7

8}, Kirti Shetty⁹, Benjamin Tran⁴, Megumi Yokomizo⁵, Preeti Ahuja^{5

6}, Yazhen Zhu^{6

10}, Hsian-Rong Tseng^{6

10}, Denise R Aberle^{5

6}, Vatche G Agopian^{4

6}, Steven-Huy B Han¹¹, Samuel W French^{2

6}, Steven M Dubinett^{2

6

10

11

12}, Xianghong Jasmine Zhou^{13

14

15}, Wing Hung Wong^{16

17}

Affiliations

¹ Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA. shuoli@ucla.edu.
² Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
³ EarlyDiagnostics Inc., Los Angeles, CA, 90095, USA.
⁴ Department of Surgery, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
⁵ Department of Radiological Sciences, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
⁶ Jonsson Comprehensive Cancer Center, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
⁷ Institute for Bioelectronic Medicine, Divisions of Gastroenterology and Hepatology, Department of Medicine, Feinstein Institutes for Medical Research, Northwell Health, Manhasset, NY, 11030, USA.
⁸ Department of Surgery, George Washington University, Washington, DC, 20037, USA.
⁹ Department of Gastroenterology and Hepatology, the University of Maryland, School of Medicine, Baltimore, MD, 21201, USA.
¹⁰ Department of Molecular and Medical Pharmacology, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
¹¹ Department of Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
¹² VA Greater Los Angeles Health Care System, Los Angeles, CA, 90073, USA.
¹³ Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA. XJZhou@mednet.ucla.edu.
¹⁴ Jonsson Comprehensive Cancer Center, University of California at Los Angeles, Los Angeles, CA, 90095, USA. XJZhou@mednet.ucla.edu.
¹⁵ Institute for Quantitative and Computational Biosciences, University of California at Los Angeles, Los Angeles, CA, 90095, USA. XJZhou@mednet.ucla.edu.
¹⁶ Department of Statistics, Stanford University, Stanford, CA, 94305, USA. whwong@stanford.edu.
¹⁷ Department of Biomedical Data Science, Stanford University, Stanford, CA, 94305, USA. whwong@stanford.edu.

PMID: 41736096
DOI: 10.1186/s13059-026-04006-0

Free article

Reducing demographic bias in biomedical machine learning for cancer detection using cfDNA methylation

Shuo Li et al. Genome Biol. 2026.

Free article

. 2026 Feb 25.

doi: 10.1186/s13059-026-04006-0. Online ahead of print.

Authors

Affiliations

¹ Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA. shuoli@ucla.edu.
² Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
³ EarlyDiagnostics Inc., Los Angeles, CA, 90095, USA.
⁴ Department of Surgery, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
⁵ Department of Radiological Sciences, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
⁶ Jonsson Comprehensive Cancer Center, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
⁷ Institute for Bioelectronic Medicine, Divisions of Gastroenterology and Hepatology, Department of Medicine, Feinstein Institutes for Medical Research, Northwell Health, Manhasset, NY, 11030, USA.
⁸ Department of Surgery, George Washington University, Washington, DC, 20037, USA.
⁹ Department of Gastroenterology and Hepatology, the University of Maryland, School of Medicine, Baltimore, MD, 21201, USA.
¹⁰ Department of Molecular and Medical Pharmacology, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
¹¹ Department of Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA.
¹² VA Greater Los Angeles Health Care System, Los Angeles, CA, 90073, USA.
¹³ Department of Pathology and Laboratory Medicine, David Geffen School of Medicine, University of California at Los Angeles, Los Angeles, CA, 90095, USA. XJZhou@mednet.ucla.edu.
¹⁴ Jonsson Comprehensive Cancer Center, University of California at Los Angeles, Los Angeles, CA, 90095, USA. XJZhou@mednet.ucla.edu.
¹⁵ Institute for Quantitative and Computational Biosciences, University of California at Los Angeles, Los Angeles, CA, 90095, USA. XJZhou@mednet.ucla.edu.
¹⁶ Department of Statistics, Stanford University, Stanford, CA, 94305, USA. whwong@stanford.edu.
¹⁷ Department of Biomedical Data Science, Stanford University, Stanford, CA, 94305, USA. whwong@stanford.edu.

PMID: 41736096
DOI: 10.1186/s13059-026-04006-0

Abstract

Background: Machine learning models in biomedical research are often hindered by demographic imbalances in clinical datasets, leading to biased predictions that disadvantage minority populations. Existing bias-correction methods face limitations in handling the heterogeneity of biomedical data and the complexity of demographic influences.

Results: We present DeBias, a computational framework for mitigating demographic biases in high-dimensional biomedical datasets. DeBias identifies and removes bias-associated subspaces from the feature space using control samples, enabling global correction of demographic distortions while preserving disease-specific signals. To evaluate its effectiveness, we apply DeBias to cell-free DNA methylation data for cancer detection. DeBias achieves a significant reduction in the number of features exhibiting demographic bias and outperforms existing methods in improving cancer detection performance for minority populations. Performance gains are validated in independent cohorts, highlighting the robustness of the approach.

Conclusions: DeBias offers an effective and generalizable strategy for correcting demographic biases in biomedical machine learning. It represents a step toward more equitable machine learning models that can deliver reliable and unbiased predictions across diverse patient populations.

Keywords: Bias correction; Cancer detection; CfDNA methylation; Demographic bias; Machine learning.

PubMed Disclaimer

Conflict of interest statement

Declarations. Ethics approval and consent to participate: The institutional review board (IRB) of the University of California at Los Angeles approved this study (IRB#19-000618, IRB#19-000230, IRB#19-001488, IRB#16-000659), and our research complies with all relevant ethical regulations. All participants gave their written informed consent. Consent for publication: Not applicable. Competing interests: X.J.Z., W.H.W., and W.L. are co-founders and board members of EarlyDiagnostics, Inc. X.J.Z. has an executive leadership position at EarlyDiagnostics, Inc. M.L.S, X.N., and C.-C.L. are employees of EarlyDiagnostics,Inc and S.M.D. was a scientific advisor to EarlyDiagnostics, Inc. X.J.Z., W.L., and W.H.W. are stockholders of EarlyDiagnostics, Inc. M.L.S, W.Z., S.L., C.-C.L., Y.Z., X.N. have stock options with EarlyDiagnostics, Inc. S.L., W.L., W.Z., and Y.Z. are consultants for EarlyDiagnostics, Inc. X.J.Z., C.-C.L. X.N., M.L.S, and W.Z. are inventors on a patent application submitted by the Regents of the University of California and EarlyDiagnostics, Inc. (Patent No. WO2023283591A2). The other authors have no competing interests to declare.

References

1. Li S, Noor ZS, Zeng W, Stackpole ML, Ni X, Zhou Y, et al. Sensitive detection of tumor mutations from blood and its application to immunotherapy prognosis. Nat Commun. 2021;12(1):4172. https://doi.org/10.1038/s41467-021-24457-2. PMID: 34234141; PMCID: PMC8263778.
1. Zviran A, Schulman RC, Shah M, Hill STK, Deochand S, Khamnei CC, et al. Genome-wide cell-free DNA mutational integration enables ultra-sensitive cancer monitoring. Nat Med. 2020;26(7):1114–24. https://doi.org/10.1038/s41591-020-0915-3. Epub 2020 Jun 1. PMID: 32483360; PMCID: PMC8108131.
1. Li S, Zeng W, Ni X, Zhou Y, Stackpole ML, Noor ZS, et al. CfTrack: a method of exome-wide mutation analysis of cell-free DNA to simultaneously monitor the full spectrum of cancer treatment outcomes including MRD, recurrence, and evolution. Clin Cancer Res. 2022;28(9):1841–53. https://doi.org/10.1158/1078-0432.CCR-21-1242. PMID: 35149536; PMCID: PMC9126584.
1. Zeng W, Liu CC, Li S, Zhou Y, Stackpole ML, Xiao Y, et al. Toward the simultaneous detection of multiple diseases with a highly cost-effective cell-free DNA methylome test. Datasets. European Genome-phenome Archive. https://www.ega-archive.org/studies/EGAS00001008125. 2025
1. Gao Q, Lin YP, Li BS, Wang GQ, Dong LQ, Shen BY, et al. Unintrusive multi-cancer detection by circulating cell-free DNA methylation sequencing (THUNDER): development and independent validation studies. Ann Oncol. 2023;34(5):486–95. https://doi.org/10.1016/j.annonc.2023.02.010. Epub 2023 Feb 26. PMID: 36849097.

Grants and funding

LinkOut - more resources

Full Text Sources
- BioMed Central

Save citation to file

Email citation

Add to Collections

Add to My Bibliography

Your saved search

Create a file for external citation management software

Your RSS Feed

Reducing demographic bias in biomedical machine learning for cancer detection using cfDNA methylation

Affiliations

Reducing demographic bias in biomedical machine learning for cancer detection using cfDNA methylation

Authors

Affiliations

Abstract

Conflict of interest statement

References

Grants and funding

LinkOut - more resources

Full Text Sources