High Efficiency Long Insert Mate Pair Library Preparation for NGS Platforms

Next generation DNA sequencing (NGS) instruments can produce 10¹⁴ bases per run. However, de novo genome assembly is compromised by three serious technical limitations; the length of reads is short, the size of the sequenced segments is also short, and the methods utilize random sequencing and assembly strategies. Millions of short random sequences are assembled into contigs by comparing the similarities. This process cannot resolve repeats or segmentally duplicated sequences, which results in large missing segments of eukaryotic genomes. This can lead to gaps, misassembled contigs, and/or collapsed repeats, leaving these regions to be resolved and finished manually, if at all. These limitations make it difficult and expensive to assemble genomes de novo.  We are developing new tools for NGS incorporating long-span true mate-pair library construction technology and software for de novo sequence assembly.  Our strategy for high efficiency mate-pair library construction involves the use of several methods to prevent chimera formation, and we have incorporated Chimera Codes™ that allow one to distinguish true mate pairs from chimeras for de novo sequence assembly.  Ion Torrent sequence data from an 8kb E. coli mate-pair library confirmed our expected results of 90% assembled reads (114,869 of 126,930) showing consistent alignment across the entire genome.  Long-span, mate-pair sequences with long connectivity and high physical coverage of a genome can be used for de novo assembly and to detect genomic variation more efficiently. Here we present data demonstrating feasibility and efficiency of these methods.