Fragment bit scores and classification rule. A) The figure shows the distribution of the fragment bit scores at different fragment lengths. The separation between the qnr fragments (light blue) and non-qnr fragments (light red) increase for longer fragment lengths. The solid blue and red lines show the average bit scores for qnr and non-qnr fragments, with their 99th and 1st percentiles in grey dashed lines above and below, respectively. The thick dashed line in black shows the classification function with the optimized parameters K=0.778, M=-7.89, D=150.64 [see Additional file 1: Figure S1, Additional file 2: Figure S2, Additional file 3: Figure S3, Additional file 4: Figure S4, Additional file 5: Figure S5 for plots corresponding to each separate class of qnr]. B) The bit scores when compared to the hidden Markov model for 33 amino acid long fragments, corresponding to the approximately 100 nucleotides long sequence reads common in next-generation sequencing technologies. At this fragment length, the qnr fragments (blue) are clearly separated from the non-qnr (red) with only a small overlap.
Boulund et al. BMC Genomics 2012 13:695 doi:10.1186/1471-2164-13-695