Comparison of coverage homogeneity and GC bias between different techniques. (A) The distribution of per-base depths was calculated (with only uniquely aligned reads) for our heat-inactivation protocol using either Covaris fragmentation (black) or Bandelin fragmentation (red), and is comparable to the standard library preparation, in which Covaris fragmentation was used in combination with SPRI cleanups (blue). (B) The GC bias is low for all compared techniques, as depicted on the right, with a slightly larger bias for the heat-inactivation protocols (using a mean depth of 200 bp bins, LOESS smooth with span = 0.3).
Wilkening et al. BMC Genomics 2013 14:90 doi:10.1186/1471-2164-14-90