Figure 1. The xGen Off-Target QC Method expands target region. In this example, a panel of hybridization probes created to capture fragments of human chromosome 21 were screened for off-target hybridization using 2 different QC methods. The repeat annotation (RA) method was compared to the xGen Off-Target QC Method in their ability to refine a custom hybridization panel that maximized the target region without increasing the amount of off-target hits. As shown in the Venn diagram, 63.9% of the example target space passed both off-target QC methods. Notably, 35.8% of the example target space passed the xGen Off-Target QC Method, whereas 0.3% of the example target space passed only the RA-based off-target QC method.

Figure 2. The xGen Off-Target QC Method does not increase the amount of off-target reads. (A) The xGen Off-Target QC Method adds design space with minimal impact on the on-target rate compared to the repeat annotation based off-target QC method. (B) The xGen Off-Target QC Method minimizes the number of off-target reads contributed by the additional probes compared to the repeat annotation based off-target QC method (C-D). The xGen Off-Target QC Method adds additional design space without increasing the percentage of ambiguity and zero coverage targets. The blue X’s represent experimental replicates, and the red triangles represent the average of three replicates.

Figure 3. Example of uniform sequence coverage and high on-target rates with xGen Custom Hyb Panels. DNA libraries created from 100ng of one (1) human genomic DNA (Coriell NA12878), in triplicate, run on 3 lots on 1 day and were enriched for an 89 kb target region, using an xGen Custom Hyb Panel—Production. The enriched libraries were sequenced on an Illumina NextSeq® 550 instrument and subsampled to 1M reads. The data show (A) consistent mean target coverage for 2 synthesis lots of 999 probes (a subset of the full experimental data shown here). The mean of triplicates for each synthesis lot are plotted. On-target rates (top) and coverage (bottom) for 3 synthesis lots were calculated with Picard metrics (Broad Institute), and plotted as mean of triplicates (B).

Figure 4. xGen Custom Hyb Probes can be added to xGen Custom Hyb Panels to expand the target regions of interest (ROIs). To customize an existing hyb panel, xGen Custom Hyb Probes can be added to the existing panels. Libraries were prepared from 100ng of one (1) human genomic DNA sample (Coriell NA12878), in triplicate, on one (1) day and captured using either an xGen Custom Hyb Panel (5001 probes targeting 33 genes, covering a 600 kb region) or the same panel spiked with additional xGen Custom Hyb Probes targeting 32 additional genes (89 kb). (A) Coverage of xGen Custom Hyb Panel, with and without the spike-in of xGen Custom Hyb Probes, demonstrates that adding additional probes does not impact coverage of the main panel. (B) Representative Integrated Genome Viewer (IGV) screenshots are shown for PIK3CA (targeted in main xGen Custom Hyb Panel) and KRAS (additional content from spike-in xGen Custom Hyb Probes). Probe regions are indicated in blue.

Figure 5. xGen Custom Hyb Panels are functionally tested prior to delivery. 100 ng of two (2) human genomic DNA samples (Coriell NA12878 and NA24385) sample was used to prepare NGS libraries. 500 ng of these libraries (n=5) were manually captured in triplicate, on one (1) day with a custom lung cancer panel (208 kb). Data from this experiment were used to generate a functional NGS report (left). The same custom lung cancer panel was then captured by the customer on a Biomek i7 (n=14, right). NGS functional data is pulled from Picard HS Metrics. All samples achieved 100X coverage or deeper.