Real-time selective sequencing using nanopore technology

Abstract

The Oxford Nanopore Technologies MinION sequencer enables the selection of specific DNA molecules for sequencing by reversing the driving voltage across individual nanopores. To directly select molecules for sequencing, we used dynamic time warping to match reads to reference sequences. We demonstrate our open-source Read Until software in real-time selective sequencing of regions within small genomes, individual amplicon enrichment and normalization of an amplicon set.

Figure 1

Targeted sequencing of specific regions of the lambda genome by direct selection using “Read Until”. (a) shows selective enrichment of the lambda genome in two 5kb regions (10-15 kb and 30-35 kb) sequencing with SQK5 chemistry (30b/s). Read Until is only applied to even numbered channels. (b) repeats this experiment on simulated reads under ideal conditions. Note the more consistent read lengths. (c) shows selective sequencing of lambda using SQK6 chemistry (70 b/s) enriching at 10-15 kb and 35-40 kb. Read Until is only applied to even numbered channels. (d) shows selective sequencing on all channels of one 10 kb region (15-25 kb). Violin plots show 2D read length for each library except (b) which is a template only simulation.

Figure 2

Selective sequencing of specific individual amplicons from a library using “Read Until”. (a) shows coverage plots for amplicons sequenced without Read Until applied. (b) shows selective sequencing of odd numbered amplicons. 2D reads are absent from even numbered amplicons. Template reads can be seen for all amplicons. (c) shows the inverse relationship where even numbered amplicons are selected for sequencing. The peaks in template coverage seen are the short reads being rejected rapidly by read until. Supplementary Fig. 8 repeats (b) showing this same phenomenon on odd numbered amplicons. Note that the lower quality of complement reads results in lower coverage when mapped to a reference and thus 2D coverage is often higher than either template or complement.

Figure 3

Normalization of amplicons in a library using “Read Until”. (a) shows coverage over each of the 11 amplicons without “Read Until” (top pane) and with “Read Until” applied (bottom pane) trying to normalize coverage. Gray - 2D reads, Orange - complement, Blue - Template. (b) shows the cumulative coverage for three representative amplicons (3,4,6) in 1 minute intervals. The black line is predicted 2D coverage for each amplicon calculated by fitting to the first 15 minutes of data. (c) shows the observed 2D coverage after 48 minutes (top pane), the total coverage predicted for each amplicon based on the first 15 minutes of sequencing (middle pane) and the projected depth if the reads were allowed to accumulate without “Read Until” until each amplicon exceeded the minimum coverage threshold, which would take approximately 105 minutes (bottom pane).