Genome Mapping Assisted De Novo Sequencing : Goliath Helps David

The ever decreasing cost per base of DNA sequencing together with the increasing output from the currently available sequencing equipment has resulted in a steep increase in the interest to sequence more genomes from which no reference sequence is known, nor from closely related species which would allow for reference mapping. Current common practice consists of brute force shotgun sequencing covering the unknown genome 50 - 100 times, followed by an attempt to reconstitute the genome by assembling all the obtained short reads. Because of the presence of repetitive sequences and particular sequences that are less efficiently captured by this classical approach (e.g. GC rich stretches in Illumina sequencing) such an approach usually results in quit a few unordered contigs, the number and average length depending on the sequencing depth and the genome complexity from the organism of interest. The longer reads from Pacific Biotechnology can already reduce the number of contigs and hence increase the contig length to some extend as those longer reads can span larger repetitive regions in the unknown genome. However, for repetitive elements that exceed the read length of a few kbp such as transposable elements in bacterial or plant genomes, alternative technologies are required that take advantage of much longer DNA molecules. In the genome mapping approach such as developed by BioNano Genomics, high molecular weight DNA is isolated from nuclei inbedded in agarose plugs. The DNA is labeled by incorporation of a fluorescent label based on the distribution of a particular recognition sequence for a nicking enzyme. Fluorescently labeled DNA is separated in a linear fashion in nanochannels and the label distribution over the long molecules is determined and converted to a genomic map of the different molecules. Short read based contigs are converted to in silico genomic maps, which are compared to the genomic maps that are experimentally obtained. Contigs that map to the same molecule can now be ordered and the extend of overlap or the size of the gap in between can be determined, hence substantially improving the quality of the assembly. In this poster/presentation we will present the results we have obtained by applying the BioNano Genomics genome mapping technology to the de novo assembly of some of the genomes analyzed so far.