Influence of Different Next-Generation Plataforms on GC-Bias - Corynebacterium pseudotuberculosis

The emergence of high throughput sequencing (HTS) platforms increased the amount of data making feasible to obtaining complete genomes. Despite the throughput produced by these platforms, the different genomic coverage in the regions of the genome can be related to GC content. This GC bias may affect genomic/transcriptomic analysis based on de novo or reference approach. In addition, the ways to evaluate the GC bias should be fit to the relationship GC vs coverage relationship with different dispersion profiles: linear, quadratic, etc.

Thus, this work proposes the use of Pearson's Correlation Coefficient (r) to analyze the correlation between GC content and coverage. To evaluate the bias, we used the data of Corynebacterium pseudotuberculosis obtained from different sequencing technologies to identify if the CG bias is related to used platforms.

Our analysis identified a non-linear and quadratic association for all sequenced strains of C. pseudotuberculosis by SOLiD platform, a linear associations for all strains sequenced by the Illumina platform, which had the highest intensity of linear association. The Ion Torrent PGM platform had a weak linear association when compared to the 454 and Illumina platforms.

Thus, the association of the different platforms with GC bias was notorious and the Pearson coefficient was effective to measure and identify different patterns of this GC bias, allowing to identify directly and inversely proportional relationships between GC and coverage.