Abstract

Droplet Barcoding for High Fidelity Single Molecule Sequencing

Naiwen Cui, Graduate Student from Weitzlab Harvard University, Harvard University

The need to accurately identify minor variants in heterogeneous mixtures of nucleotides arises in many contexts such as research, clinical diagnostics, and forensic analysis. Examples include detecting drug resistance mutations in tumor biopsies, criminal forensic biology, etc.. Next-generation sequencing has enabled sensitive detection and rapid genotyping of low-proportion contributor variants using PCR to selectively amplify targeted region prior to sequencing. However this process generates artifact sequences, which dramatically increases the background noise and thus makes it extremely difficult for identification of minor variant from a heterogeneous sample. Here we present the High Fidelity Sequencing method, which allows isolation of individual DNA template along with one Hydrogel Barcoded Bead to uniquely tag the PCR amplicon products inside every drop prior to deep sequencing. Information from each drop is then analyzed individually based on their barcode sequence. Combining with novel algorithmic processing, we significantly reduce the background noise by over 1000 fold to yield virtually error-free single template sequence data. We successfully applied this method to detect <1% minor contributor alleles with no false positives in a 5-person mixture sample as well as on mock forensic samples. This method has a great potential to revolutionize the field of single molecule sequencing.