Fig. 1

The main findings and performance of double-batched sequencing (DoBSeq) in the explorative and validation cohorts. Upper panel: At the top; a timeline showing how all patients in the explorative cohort had neonatal blood spots taken at birth, followed by a presymptomatic phase prior to a cancer diagnosis (equivalent to neonatal blood sample age), after which they underwent whole genome sequencing which identified several loss-of-function or reported pathogenic variants (LoF/P). A Jitter plot showing LoF/P variants detected in the 10-column batches, plotted with variant allele frequency (VAF) on the y-axis. Blue and labeled dots represent true positive variants while red and unlabeled dots represent false positives. The gray dotted line represents the theoretically expected VAF of 5% for non-mosaic heterozygous variants (1 of 20 alleles). B Jitter plot shows the same as A only for row batches. C Doubly detected LoF/P variants are pinned to a specific patient in a matrix where each intersection represents one sample/patient. Dots represent variants identified by DoBSeq. Teal gene names indicate true positives (also found on WGS) and red gene names show false positives (not found on WGS). Higher transparency of dots/gene names indicates lower confidence. Lower panel: D–F are identical to A–C, only showing data and results from the validation cohort (VC). * The number of LoF/P variants found on whole genome sequencing refers to SNVs and indels within the exonic regions covered by the panel used in the DoBSeq matrices. See the “Methods” section for further details