Figure 2

Effects of normalization and 454 sequencing chemistry on read length and isotig length. (A) Titanium sequencing chemistry (grey, black) generally results in longer read lengths when compared with FLX chemistry (white). However, the normalized sample run with Titanium chemistry (black) had shorter read lengths than the non-normalized sample (grey). This result is likely due to a technical error in that particular sequencing run, since a 1/8 plate run of the same sample showed a read length distribution comparable to that of the non-normalized sample (Additional file 1). (B) Isotig length distributions from assemblies of Titanium-sequenced data. The longest isotig per isogroup is shown. The number of bases in the non-normalized (grey) and normalized (black) samples has been equalized to eliminate possible bias due to the greater number and length of reads obtained from the run of the normalized sample (see (A)). The isotigs generated from the normalized cDNA tended to be shorter than those produced by the non-normalized cDNA (see also Table 2). Pooling all FLX and Titanium reads generates an assembly with more, longer isotigs (blue).