Low-Cost Genome-Scale Phasing with Barcode-Linked Sequencing.

Complete comprehension of clinically relevant variation among human genomes is likely only to come from sequencing platforms that are cost-efficient, and which feature both accurate base calling and long-range DNA phasing capability. The NGS revolution has struggled to meet the latter of these needs. Here we describe a protocol to address this limitation by preserving the molecular origin of short sequencing reads with an insignificant increase to sequencing costs. Whole haplotype-resolved genomes with megabase-scale phase blocks can be obtained with this method; offering researchers a unique opportunity to tackle the hurdles of de novo sequencing without being limited by a lack of resources.

High throughput barcoding method for genome-scale phasing.

The future of human genomics is one that seeks to resolve the entirety of genetic variation through sequencing. The prospect of utilizing genomics for medical purposes require cost-efficient and accurate base calling, long-range haplotyping capability, and reliable calling of structural variants. Short-read sequencing has lead the development towards such a future but has struggled to meet the latter two of these needs. To address this limitation, we developed a technology that preserves the molecular origin of short sequencing reads, with an insignificant increase to sequencing costs. We demonstrate a novel library preparation method for high throughput barcoding of short reads where millions of random barcodes can be used to reconstruct megabase-scale phase blocks.

Thermophoresis-Controlled Size-Dependent DNA Translocation through an Array of Nanopores.

Large arrays of nanopores can be used for high-throughput biomolecule translocation with applications toward size discrimination and sorting at the single-molecule level. In this paper, we propose to discriminate DNA length by the capture rate of the molecules to an array of relatively large nanopores (50-130 nm) by introducing a thermal gradient by laser illumination in front of the pores balancing the force from an external electric field. Nanopore arrays defined by photolithography were batch processed using standard silicon technology in combination with electrochemical etching. Parallel translocation of single, fluorophore-labeled dsDNA strands is recorded by imaging the array with a fast CMOS camera. The experimental data show that the capture rates of DNA molecules decrease with increasing DNA length due to the thermophoretic effect of the molecules. It is shown that the translocation can be completely turned off for the longer molecule using an appropriate bias, thus allowing a size discrimination of the DNA translocation through the nanopores. A derived analytical model correctly predicts the observed capture rate. Our results demonstrate that by combining a thermal and a potential gradient at the nanopores, such large nanopore arrays can potentially be used as a low-cost, high-throughput platform for molecule sensing and sorting.

Droplet Barcode Sequencing for targeted linked-read haplotyping of single DNA molecules.

Data produced with short-read sequencing technologies result in ambiguous haplotyping and a limited capacity to investigate the full repertoire of biologically relevant forms of genetic variation. The notion of haplotype-resolved sequencing data has recently gained traction to reduce this unwanted ambiguity and enable exploration of other forms of genetic variation; beyond studies of just nucleotide polymorphisms, such as compound heterozygosity and structural variations. Here we describe Droplet Barcode Sequencing, a novel approach for creating linked-read sequencing libraries by uniquely barcoding the information within single DNA molecules in emulsion droplets, without the aid of specialty reagents or microfluidic devices. Barcode generation and template amplification is performed simultaneously in a single enzymatic reaction, greatly simplifying the workflow and minimizing assay costs compared to alternative approaches. The method has been applied to phase multiple loci targeting all exons of the highly variable Human Leukocyte Antigen A (HLA-A) gene, with DNA from eight individuals present in the same assay. Barcode-based clustering of sequencing reads confirmed analysis of over 2000 independently assayed template molecules, with an average of 753 reads in support of called polymorphisms. Our results show unequivocal characterization of all alleles present, validated by correspondence against confirmed HLA database entries and haplotyping results from previous studies.

Phasing of single DNA molecules by massively parallel barcoding.

High-throughput sequencing platforms mainly produce short-read data, resulting in a loss of phasing information for many of the genetic variants analysed. For certain applications, it is vital to know which variant alleles are connected to each individual DNA molecule. Here we demonstrate a method for massively parallel barcoding and phasing of single DNA molecules. First, a primer library with millions of uniquely barcoded beads is generated. When compartmentalized with single DNA molecules, the beads can be used to amplify and tag any target sequences of interest, enabling coupling of the biological information from multiple loci. We apply the assay to bacterial 16S sequencing and up to 94% of the hypothesized phasing events are shown to originate from single molecules. The method enables use of widely available short-read-sequencing platforms to study long single molecules within a complex sample, without losing phase information.