SituSeq: an offline protocol for rapid and remote Nanopore 16S rRNA amplicon sequence analysis

Abstract

Microbiome analysis through 16S rRNA gene sequencing is a crucial tool for understanding the microbial ecology of any habitat or ecosystem. However, workflows require large equipment, stable internet, and extensive computing power such that most of the work is performed far away from sample collection in both space and time. Performing amplicon sequencing and analysis at sample collection would have positive implications in many instances including remote fieldwork and point-of-care medical diagnoses. Here we present SituSeq, an offline and portable workflow for the sequencing and analysis of 16S rRNA gene amplicons using Nanopore sequencing and a standard laptop computer. SituSeq was validated by comparing Nanopore 16S rRNA gene amplicons, Illumina 16S rRNA gene amplicons, and Illumina metagenomes, sequenced using the same environmental DNA. Comparisons revealed consistent community composition, ecological trends, and sequence identity across platforms. Correlation between the abundance of taxa in each taxonomic level in Illumina and Nanopore data sets was high (Pearson's r > 0.9), and over 70% of Illumina 16S rRNA gene sequences matched a Nanopore sequence with greater than 97% sequence identity. On board a research vessel on the open ocean, SituSeq was used to analyze amplicon sequences from deep sea sediments less than 2 h after sequencing, and 8 h after sample collection. The rapidly available results informed decisions about subsequent sampling in near real-time while the offshore expedition was still underway. SituSeq is a portable and user-friendly workflow that helps to bring the power of microbial genomics and diagnostics to many more researchers and situations.

Fig. 1. Research expedition on the Scotian slope offshore Nova Scotia.

A Sites sampled in the Northwest Atlantic Ocean. B ROV footage from the background site “175NW”. C ROV footage from the hydrocarbon seep site “Purple Haze”. D Image of the 31 cm long core taken from site “175NW”. E Image of the 36 cm long core taken from site “Purple Haze”. F A Nanopore MinION sequencing run performed while at sea during the sampling expedition.

Fig. 2. SituSeq bioinformatics workflow.

The preprocessing script contains quality control steps that remove primers and filter sequences outside the specified length parameters. The Stream 1 workflow uses the assignTaxonomy function from dada2 to assign taxonomy to all Nanopore sequences. The Stream 2 workflow performs a BLASTn search of the Nanopore sequences as queries against a custom database.

Fig. 3. Relative abundance of most abundant phyla.

A Abundance of phyla across samples from Nanopore data set. B Abundance of phyla across samples from Illumina data set. “Unknown” represents the sequences not identified at the phylum level and “Other” represents the less abundant phyla.

Fig. 4. Comparison of Nanopore and Illumina amplicon sequencing.

A Correlation between abundance of phyla sequenced with Nanopore (y-axis), and Illumina (x-axis) technology (Pearson’s r = 0.905). Abundant phyla are colored, “Unknown” represents the sequences not identified at the phylum level, and “Other” represents the less abundant phyla. The black dashed line shows the linear relationship between Nanopore and Illumina abundances, and the red line shows a 1:1 ratio. The axes have been square root transformed. B The ratio of Illumina abundance to Nanopore abundance of select phyla. Ratios for individual samples are overlaid on the boxplots. The red line shows a 1:1 ratio. The same comparisons at other taxonomic levels can be found in Figs. S1–3.

Fig. 5. NMDS plots with combined Nanopore and Illumina data sets.

The same NMDS plots based on Bray-Curtis dissimilarity are shown with samples (points) colored and sized based on different parameters. A Sample color indicates sequencing technology. B Samples colored on location, with size proportional to depth below the sediment surface. Stress = 0.1.

Fig. 6. Comparison of relative abundance of a selection of differentially abundant phyla between Nanopore amplicon, Illumina amplicon, and Illumina metagenome data sets.

Note the different scales of the y-axes. See also Table S3.

Fig. 7. Combined relative abundance and distribution of 21 taxa of interest in each sample.

SituSeq Stream 2 was used to conduct a BLAST search against a custom database of species indicative of hydrocarbon seepage (Data S2). The shaded gray area shows the depths of the samples that were sequenced for each core.