Fig. 2

Generating a targeted sequencing library for hybrid capture of LB mutations. We generated a candidate panel consisting of probes targeting the ~ 7000 highest ranked LB mutation loci. a We binned genes represented by candidate mutations into 10 groups based on length and show the distribution in number of mutations. Gene length correlated with the number of mutations on the panel (Pearson’s correlation = 0.20, p = 6.03e-39). b We employed a distance standardization to mutation hyperplane distances to increase gene diversity on the panel. After standardization the correlation between gene length and number of mutations decreased significantly (Pearson’s correlation = 0.05, p = 0.0015). c We plotted the hyperplane distances of retained mutations after standardization against the log mutation rank. Mutations are labeled as coding (green) or non-coding (grey). The top 5 coding mutations with their corresponding genes are labeled. d We show a table of panel mutation consequence types and counts, colored by impact severity (red = high, orange = moderate, yellow = low, blue = modifier)