Back

Cross-dataset pan-cancer detection: Correlating cell-free DNA fragment coverage with open chromatin sites across cell types

Olsen, L. R.; Odinokov, D.; Holsting, J. Q.; Kondrup, K.; Iisager, L.; Rusan, M.; Buus, S.; Laursen, B. E.; Borre, M.; Jochumsen, M. R.; Bouchelouche, K.; Frydendahl, A.; Rasmussen, M. H.; Henriksen, T. V.; Nesic, M.; Demuth, C.; Lindskrog, S. V.; Nordentoft, I.; Lamy, P.; Therkildsen, C.; Dyrskjot, L.; Sorensen, K. D.; Andersen, C. L.; Skanderup, A. J.; Besenbacher, S.

2024-11-28 oncology
10.1101/2024.11.26.24317971
Show abstract

The fragmentation patterns of whole genome sequenced cell-free DNA are promising features for tumor-agnostic cancer detection. However, systematic biases challenge their cross-cohort generalization. We introduce LIONHEART, a novel, open source cancer detection method specifically optimized to generalize across datasets. The method correlates bias-corrected cfDNA fragment coverage across the genome with the locations of accessible chromatin regions from 898 cell and tissue type features. We use these correlations to detect changes in the cell-free DNA cell type composition caused by cancer. We test LIONHEART on nine datasets and fourteen cancer types (1106 non-cancer controls, 1449 cancers) obtained from different studies and show that it can distinguish cancer samples from non-cancer controls across cohorts with ROC AUC scores ranging from 0.62-0.95 (mean = 0.83, std = 0.12). We further validate the method on an external dataset, achieving a ROC AUC of 0.917.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

1
Nature Communications
based on 483 papers
Top 0.8%
21.1%
2
npj Precision Oncology
based on 14 papers
Top 0.1%
11.9%
3
Scientific Reports
based on 701 papers
Top 13%
10.9%
4
PLOS Computational Biology
based on 141 papers
Top 3%
4.8%
5
PLOS ONE
based on 1737 papers
Top 75%
3.2%
50% of probability mass above
6
Nature Genetics
based on 72 papers
Top 4%
3.2%
7
Clinical Cancer Research
based on 22 papers
Top 2%
3.0%
8
JCO Clinical Cancer Informatics
based on 14 papers
Top 1%
2.6%
9
Nature
based on 58 papers
Top 3%
2.5%
10
iScience
based on 74 papers
Top 2%
2.5%
11
eLife
based on 262 papers
Top 11%
2.5%
12
Cancers
based on 57 papers
Top 5%
1.7%
13
Proceedings of the National Academy of Sciences
based on 100 papers
Top 8%
1.4%
14
Nature Medicine
based on 88 papers
Top 9%
1.4%
15
EMBO Molecular Medicine
based on 15 papers
Top 1%
1.3%
16
International Journal of Radiation Oncology*Biology*Physics
based on 13 papers
Top 2%
0.9%
17
JCO Precision Oncology
based on 11 papers
Top 2%
0.9%
18
Frontiers in Oncology
based on 34 papers
Top 5%
0.9%
19
Journal for ImmunoTherapy of Cancer
based on 14 papers
Top 2%
0.9%
20
Cancer Epidemiology, Biomarkers & Prevention
based on 14 papers
Top 3%
0.9%
21
Leukemia
based on 11 papers
Top 1%
0.9%
22
Communications Biology
based on 36 papers
Top 4%
0.9%
23
Breast Cancer Research
based on 11 papers
Top 2%
0.7%
24
Briefings in Bioinformatics
based on 11 papers
Top 0.7%
0.7%
25
Science Translational Medicine
based on 40 papers
Top 5%
0.7%
26
Cancer Medicine
based on 17 papers
Top 4%
0.7%