highly significantp = 5×10 -36
The read accuracy improvement of using ORP over single reads, revealed as a gap between their respective error profiles, was most prominent at Q10 and diminished with higher Q-score filtering; the advantage at Q30 was small yet still highly significant (two-sample Kolmogorov-Smirnov test p = 5×10 -36 and 2×10 -11 for rabies and BCV control, respectively), as any reduction in error rate can be used to detect viral variants with greater sensitivity.