Nanopore adaptive sampling: a tool for enrichment of low abundance species in metagenomic samples

Abstract

Adaptive sampling is a method of software-controlled enrichment unique to nanopore sequencing platforms. To test its potential for enrichment of rarer species within metagenomic samples, we create a synthetic mock community and construct sequencing libraries with a range of mean read lengths. Enrichment is up to 13.87-fold for the least abundant species in the longest read length library; factoring in reduced yields from rejecting molecules the calculated efficiency raises this to 4.93-fold. Finally, we introduce a mathematical model of enrichment based on molecule length and relative abundance, whose predictions correlate strongly with mock and complex real-world microbial communities.

Keywords: Adaptive sampling; Enrichment; Metagenomics; Nanopore; ReadUntil; Sequencing.

Fig. 1

a Read length distributions from control runs. Reads were binned by length into bins of size 1000 bp. Distribution for 10.6 kbp library taken from control run in high-to-low experiment. b Violin plots (log scale) of read length distributions from control runs. Distribution for 10.6 kbp library taken from control run in high-to-low experiment. Extrema and means shown in black. c Enrichment factor against relative abundance. Each point represents a species, with the position on the x-axis indicating the original relative abundance of the species and the position on the y-axis indicating the enrichment factor obtained. d Community composition for each enrichment target during the runs

Fig. 2

a Scatterplots of enrichment vs abundance. Curves show enrichment values predicted by the model for average read lengths. b Correlation between observed enrichment values and predicted enrichment (Pearson’s r of 0.9825)

Fig. 3

a Yield of target sequences in Mb per hour during adaptive sampling (blue), control before/during (red), and control after (purple). b Yields per hour for all runs, normalised by channels used. c Yield of target sequences in Mb per hour per active channel during adaptive sampling. d Enrichment by yield values. Each experiment, except for the 1.7 kbp run, gave us increased yield when performing adaptive sampling

Fig. 4

a Distribution of read lengths during control portion and enrichment portions of 12.8 kbp run. Reads are split by species. b Proportion of target reads rejected during adaptive sampling. c Quality values of reads, split by species and TP/FN. d Average identity of mappings of first 200 bp of reads against reference genomes. The mapping to the correct genome with the highest identity was used to calculate the averages. e Coverage of target genomes by false negative reads (i.e. reads that were incorrectly ejected from the pore during adaptive sampling) during 12.8 kbp run. Image produced using the alignment visualisation software Alvis [25]

Fig. 5

Cumulative yields split by experiment channels and control channels

Fig. 6

a S. dysgalactiae assembly statistics for enriched and control channels. b Plots showing how the number of active channels varies with time. c Hourly yields from enriched/depleted channels vs control channels. d Times between consecutive target molecules on individual channels, split by enrich/deplete (channels 1–256, red) and control (channels 257–512, blue)

Fig. 7

Effect of nuclease flush on active channels and yield

Fig. 8

Enrichment of selected taxa in complex microbial community. Strains of E. coli highlighted separately. a Enrichment by composition, includes enrichment curve as predicted by model. b Enrichment by yield. c Enrichment values for each strain of E. coli by targeted strain