Table 1 F1-score (%) in dataset A experiment-1 and experiment-2 for each tested ROI strategy expressed as mean ± standard deviation (SD) and as median (minimum–maximum). ROI-8 shows the highest F1 mean in both experiments. The high standard deviation is explained by the high rate of false positive variants in the non-KIV-2B samples, which range from mean = 0 with ROI-9 (both experiments 1 and 2) to mean = 152 and mean = 164 with ROI-1 (experiments 1 and 2, respectively). See Additional file 1: Table S4 for performance metrics at the individual level

From: Resolving intra-repeat variation in medically relevant VNTRs from short-read sequencing data using the cardiovascular risk gene LPA as a model