Fig. 3

Overall performance of the VIS-NGS protocol. The VIS-NGS protocol with LM-NEO-PCR only or combined with NEO-capture followed by NGS detected 12,857 loci in 33 samples, including replicate samples. Most loci were detected in only one sample, while 1031 loci were identified in at least 2 samples (A). Loci in one sample were detected at low read coverage (reads per million bases (RPMs), presumably noise, whereas the loci identified in at least 2 samples were detected at higher read coverage (p = 0.013). Median read coverages are presented excluding outliers in both groups (B). Total read counts coverage and number of detected loci was compared between samples without VIS loci (negative controls, blue boxes), and samples with VIS loci from single pure clone (orange boxes), and from composed mixed and diluted samples having two clones (grey boxes) or four clones (yellow boxes). The VIS-NGS protocol produced significantly (p < 0.016) more reads in samples having VIS loci compared to the negative control samples (C). Horizontal dashed lines show comparison and their p values (student T-test). No significant (p > 0.12) differences were seen in the total number detected loci between samples with clones and samples without clones (negative control) (D). (E) and (F) show the number of loci detected when applying two thresholds on read coverage for VIS locus calling (loci with > 100 RPMs and with > 1000 RPMs) were compared between samples without clones (negative controls) and samples with clones. Only loci with > 1000 RPMs resulted in significant (p < 0.008) more loci in samples with clones compared to negative controls (F). No significant differences in number of loci with > 100 RPMs were seen between samples with and without clones (p > 0.09) (E). Horizontal dashed lines show comparison and their p values (student T-test).