Skip to main content
Fig. 4 | Genome Biology

Fig. 4

From: Achieving robust somatic mutation detection with deep learning models derived from reference data sets of a cancer sample

Fig. 4

Performance comparisons on the tumor/normal titration dataset. a F1-score (%) comparison for different somatic mutation callers across different coverages (10x–300) and tumor purities (5–100%). For a given coverage, results of different callers on the same tumor purity are connected for better illustration. b Robustness to tumor contamination in the matched normal: F1-score (%) when matched normal was mixed with 5% of tumor is shown versus the F1-score (%) at pure normal for 80x coverage and 5–100% tumor purities. Here, the SEQC-WGS-GT50-SpikeWGS10 trained models were used for NeuSomatic and NeuSomatic-S

Back to article page

Genome Biology

ISSN: 1474-760X

Contact us