Fig 1.
A general overview of the dsDNA Fragmentation Through Polymerization (FTP) method.
The FTP method is based on two enzymatic reactions: a DNA nicking reaction with DNase I and a strand-displacement DNA polymerization with SD DNA polymerase. As a result, multiple double-stranded DNA fragments with overlapping sequences are generated. De novo synthesized DNA is indicated in grey, and SD polymerase is indicated in red.
Fig 2.
Agarose-gel electrophoresis of gDNA fragmented by the FTP method.
gDNA of E. coli BL21 was incubated as described in Materials and Methods: without enzymes (lane 1), with SD polymerase (lane 2), with DNase I (lane 3), and with both DNase I and SD polymerase (lane 4 and 5). M1: 1 kb DNA Ladder; M2: 100 bp DNA Ladder.
Fig 3.
Comparison of the mean nucleotide compositions in the reads of FTP- and Fragmentase-generated NGS libraries.
(A) Bias plots showing mean (by replication) percentage of observed bases at each position of reads for Fragmentase (solid lines) and FTP (dotted lines) methods of fragmentation. (B) χ2 values of observed bases at each position of reads (calculated by Pearson χ2 test for given probabilities) for Fragmentase (red) and FTP (blue) methods. Given probabilities are mean probabilities for each nucleotide from position 11 to position 149. Smaller χ2 values indicate that the observed probability of bases at the given position is closer to the mean probabilities at the non-bias region and less bias is observed.
Fig 4.
Coverage uniformity evaluation.
Cumulative read coverage was visualized as Lorenz curves (A) and GC bias of the coverage was estimated as normalized coverage over GC content for both Fragmentase (red curves) and FTP (blue curves) methods. (A) The Lorenz curves show the cumulative fraction of the genome as a function of the cumulative fraction of the reads. Perfectly uniform coverage would result in a diagonal line (black). Fragmentase and FTP methods exhibit the same deviations from the diagonal as a result of biased coverage. (B) The GC bias plots show the normalized coverage as a function of GC content. The black horizontal line (normalized coverage = 1) represents an ideally uniform coverage and any divergence from it indicates either oversequencing (normalized coverage > 1) or underrepresentation (normalized coverage < 1) of the sequences of particular GC content. Both methods give similar uniformity, while FTP provides better coverage for GC reach (> 55% GC content) sequences.
Table 1.
Key averaged NGS characteristics of Fragmentase- and FTP- generated libraries.
Table 2.
The averaged assembly metrics of the NGS libraries obtained by Fragmentase and FTP methods.