Fig. 6From: SNooPer: a machine learning-based method for somatic variant identification from low-pass next-generation sequencingValidation plot. Distribution of 50 randomly selected SNVs called using SNooPer's Model 1A on the independent validation set constituted of samples obtained from 34 childhood acute lymphoblastic leukemia patients (matched normal and tumor). All selected SNVs were heterozygous with a VAF < 0.6, predicted as damaging (Sift [42 44] p-values ≤0.05) and presented a class probability >0.9. Each identified SNV was validated by targeted ultra-deep re-sequencing (>1000X). The grey line indicates the expected VAF (50%) for germline or clonal somatic heterozygous variants. Dark cyan squares, grey dots and white diamonds represent validated somatic, germline and non-validated variations respectivelyBack to article page