Using Linked-Reads to Enable Efficient de novo Diploid Assembly

Assembly of mammalian genomes using cost effective, accurate short reads is challenging. We describe a novel approach for the de novo assembly of individual mammalian genomes, requiring only small amounts (0.5-1.25 ng) of input DNA to construct a single library. We have developed a microfluidic system that allows for the high-throughput partitioning of high-molecular weight DNA. Unique barcodes are applied within each partition, allowing for the retention of long-range information using short read sequencing, creating a data type called Linked-Reads. The Supernova™ De Novo Assembler takes advantage of Linked-Reads to perform de novo diploid assembly. The process begins by constructing a De Bruijn graph using the read data. The graph is then progressively refined by using read-pair barcode information. This allows for the construction of multimegabase scaffolds. Heterozygosity within the sample, coupled with molecular bar codes, allows for the separation of these scaffolds into their distinct haplotypes, referred to as phase-blocks. The reconstruction of individual haplotypes, rather than a haploid consensus sequence, allows for a more complete, and accurate representation of the sample. We demonstrate the performance of this process on several human genomes of diverse ethnic origin. These assemblies consistently have scaffold lengths on the order of 15 Mb, with phase block lengths varying from 4 -11 Mb. As some of these samples are members of well characterized trios, we can validate accuracy of the phase information using orthogonal data. We will also show performance on a variety of other genomes including hummingbird, dog and olive fly.