Bioinformatics Final Project - Adjusting for GC bias in the read-counts

GC bias correction

Locally weighted scatterplot smoothing (lowess) regression defines the relationship between read counts and GC%

Using this fit we adjust the slope of the points in the scatterplot and get the corrected read counts

The correction is demonstrated by a reduction in the variance of the distribution (gray dots vs black dots) of the read counts across the genome.

Figure 8a: Plots above demonstrate how sample read counts can be biased towards the volume of their GC content (light blue). The correction reduces the slope and corrects for the trend (dark blue).

Figure 8b: The read distributions across the genome are visualized across the genome along with the GC% (normalized) at the same positions. Areas where low or high read counts (gray) are correlated with high or low GC% (brown), are observed to be adjusted (black).

Page updated

Report abuse